This thesis explores some important practical considerations concerning the use of species distribution models in marine conservation planning. Using geo-referenced gorgonian distribution data, together with explanatory environmental variables, predictive models have been used to map the spatial distribution of suitable gorgonian (sea fan) habitat in two study sites; Hatton Bank, in the Northeast Atlantic, and Lyme Bay on the south coast of Devon. Generalized Linear Models (GLMs), Generalized Additive Models (GAMs) and a Maximum Entropy (Maxent) model have been used to support critical investigation into important model considerations that have received inadequate attention in the marine environment. The influence of environmental data resolution on model performance has been explored with specific reference to available datasets in the nearshore and offshore environments. The transferability of deep-sea models has been similarly appraised, with recommendations as to the appropriate use of transferred models. Investigating these practical issues will allow managers to make informed decisions with respect to the best and most appropriate use of existing data. This study has also used novel approaches and investigated their suitability for marine conservation planning, including the use of model classification error in the spatial prioritisation of monitoring sites, and the adaptation of an existing presence-only modelling method to include absence data. Together, these studies contribute both practical recommendations for marine conservation planning and novel applications within the wider species distribution modelling discipline, and consider the implications of these developments for managers, to ensure the ongoing improvement and development of models to support conservation planning.

Cold-water corals, similar to tropical corals, contain a diverse and complex microbial landscape. Comprised of vital microscopic organisms (i.e. bacteria, viruses, archaea), the coral microbiome is a driving factor in the proliferation and survival of the coral host. Bacteria provide essential biological functions within coral holobionts, facilitating increased nutrient utilization and production of antimicrobial compounds. To date, few cold-water octocoral species have been analyzed to explore the diversity and abundance of their microbial associates. For this study, 23 samples of the family Anthothelidae were collected from Norfolk (n = 12) and Baltimore Canyons (n = 11) from the western Atlantic in August 2012 and May 2013. Genetic testing found that these samples comprised two Anthothela species (Anthothela grandiflora and Anthothela sp.) and a new genus. DNA was extracted and sequenced with primers targeting the V4-V5 variable region of the 16S rRNA gene using 454 pyrosequencing with GS FLX Titanium chemistry. Results demonstrated that the host genus was the primary driver of bacterial composition. The new coral genus, dominated by Alteromonadales and Pirellulales, had much higher species richness and a distinct bacterial community compared to Anthothela samples. Anthothela species had very similar bacterial communities, dominated by Oceanospirillales and Spirochaetes. Core bacterial diversity present across 90% of the Anthothela samples revealed genus-level conservation. This core included unclassified Oceanospirillales, Kiloniellales, Campylobacterales, and Spirochaeta; the functional abilities of which contribute to a nearly complete nitrogen cycle. Dominant bacterial members of the new coral genus also had functional capabilities in nitrogen cycling. Overall, many of the bacterial associates identified in this study have the potential to contribute to the acquisition and cycling of nutrients within the coral holobiont.

L’objectif de cette thèse est d’apporter de nouvelles contraintes sur l’hydrologie de l’Atlantique Nord-Est et de la Méditerranée occidentale depuis la dernière période glaciaire à partir de l’analyse de la composition isotopique du Nd (εNd) dans des échantillons d’eau de mer ainsi que des coraux profonds et des foraminifères, prélevés dans des carottes sédimentaires marines. Les changements de l’hydrologie des masses d’eau intermédiaire (LIW, MSW, AAIW et masses d’eau intermédiaire des gyres subtropical et subpolaire) ont été plus particulièrement étudiés car leur rôle sur les transferts de sels en Atlantique Nord et in fine sur l’AMOC est actuellement mal contraint. Ce travail a été mené au cours des périodes de changements hydrologiques majeurs et abrupts de l’océan qui se sont produits lors des variations climatiques rapides de la dernière période glaciaire (événements d’Heinrich et cycles de Dansgaard-Oeschger) et lors du dépôt du Sapropel S1 en Méditerranée orientale. Dans un premier volet, nous avons amélioré la couverture spatiale des valeurs d’εNd des masses d’eau de l’Atlantique Nord-Est et de la mer d’Alboran, préalable indispensable pour restituer l’hydrologie passée de ces régions avec le traceur εNd. Nous avons ensuite mis en évidence un changement majeur du schéma de circulation de la Méditerranée occidentale durant la période de dépôt du sapropel S1, marquée par une forte réduction des masses d’eau de la Méditerranée orientale (LIW) au sud de la Sardaigne au profit de celles provenant du Golfe du Lion (WIW). Ce changement hydrologique ainsi que ceux qui s’opèrent en Méditerranée depuis la dernière période glaciaire ne sont pas associés à de fortes modifications des valeurs d’εNd de la LIW de la mer d’Alboran et de la mer des Baléares, suggérant une stabilité de la signature isotopique en Nd de la MOW au cours du temps. Ceci a permis, à partir d’un enregistrement d’εNd obtenus sur des coraux profonds du Golfe de Cadix, de mettre en évidence une contribution plus importante de l’AAIW plus radiogénique et donc une pénétration plus marquée en Atlantique Nord de cette masse d’eau lors des périodes de fortes réductions de l’AMOC, liées à la déstabilisation des calottes de glace de l’Hémisphère Nord. / : The purpose of this thesis is to constrain the hydrology of the North-East Atlantic and western Mediterranean Sea since the last glacial period from neodymium isotopic composition (εNd) measured on seawater, cold water corals and foraminifera. In particular, hydrological changes of intermediate water masses (LIW, AAIW, MSW, mid-subtropical and subpolar gyre water) have been studied as their role on salt budget in North Atlantic and ultimately on AMOC are currently poorly constrained. This work has been conducted at times of major and abrupt hydrological changes that occurred during rapid climatic variations of the last glacial period (Heinrich and Dansgaard-Oeschger events) and during the last sapropel deposit (S1) in eastern Mediterranean Sea. In a first step, we have improved the spatial distribution of water masses εNd values in North-east Atlantic and Alboran Sea, what is an absolute prerequisite in order to track past hydrological changes in these areas with εNd proxy. Next, we have highlighted a major change of the western Mediterranean circulation pattern during the sapropel S1 deposit, which is marked south of Sardinia by a strong reduction of eastern-sourced water masses (LIW) in favor of western-sourced water masses (WIW). This hydrological change as well as those occurring in Mediterranean Sea since the last glacial period was not associated with strong modifications of εNd values in Alboran and Balearic Sea, suggesting a stability of Nd isotopic signature of MOW over the time. This has highlighted, from an εNd record obtained on cold water corals in the Gulf of Cadiz, an enhanced contribution of more radiogenic AAIW and therefore a stronger northward penetration in North Atlantic at times of reduced AMOC linked to iceberg discharges from Northern Hemisphere ice sheets.

Na presente dissertação eu estudei a distribuição potencial dos corais de águas profundas ao longo da margem continental brasileira e o efeito da resolução das variáveis ambientais em modelos de adequabilidade de habitat (MAH). Primeiramente foi determinada a distribuição potencial de corais de águas profundas a partir de modelos de adequabilidade de habitat de máxima entropia (MaxEnt). Os modelos foram alimentados com diversas camadas de dados SIG de variáveis ambientais e dados de ocorrência de espécies extraídos de bancos de dados e de novos registros. Foram estudados os grupos Scleractinia e Octocorallia, assim como quatro espécies de corais formadores de recifes (Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata e Enallopsammia rostrata) e duas espécies de escleractínios não formadores de recifes (Cladocora debilis e Deltocyathus spp.). O nicho ambiental ocupado pelos táxons foi representado a partir de uma análise multivariada, sendo determinada a sobreposição entre os nichos. A partir disso foi avaliada a similaridade e equivalência de nicho ambiental entre as espécies. Os modelos resultantes indicaram que Scleractinia apresenta distribuição potencial na margem continental leste e sudeste do Brasil (porção Central e Sul), enquanto que Octocorallia apresenta uma distribuição potencial mais ampla abarcando toda a margem continental brasileira, com maior adequabilidade nas margens norte e nordeste. Assim, as áreas de distribuição potencial compartilhada entre octocorais e escleractínios foram aquelas ao largo da porção leste e sudeste do Brasil, sendo estas prioritárias para futuras pesquisas. Os nichos ambientais específicos indicaram menos tolerância de Scleractinia a altas temperaturas em comparação com Octocorallia. Os octocorais ocuparam uma maior amplitude do gradiente ambiental, principalmente relacionado às condições de carbono orgânico particulado, oxigênio dissolvido e temperatura. Isto possivelmente permite aos octocorais habitar a porção norte da margem continental brasileira, onde os modelos de escleractínios indicaram baixa adequabilidade de habitat. A margem continental sudeste do Brasil teve a maior adequabilidade para as espécies de escleractínios formadores de recifes profundos. Isto é de grande valor para planos de conservação nesta área que atualmente apresenta atividades de extração de óleo e gás e que, futuramente, pode ser explotada para a extração de minério. Além disso, as espécies formadoras de recifes ocuparam nichos ambientais similares e, portanto, mudanças nas condições ambientais, como por exemplo as associadas às mudanças climáticas, podem afetar todas elas. Contudo, seus nichos não são equivalentes e, portanto, cada uma delas faz um uso diferente do espaço ambiental da margem continental brasileira. Por último, foram desenvolvidos modelos MaxEnt para Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata e Enallopsammia rostrata em duas resoluções de análise, uma de baixa (i.e., com um tamanho de pixel de aproximadamente 1000 m2) e uma de alta resolução (i.e., pixel de aproximadamente 100 m2) para determinar o efeito da resolução das variáveis ambientais utilizadas na performance e na extensão de área predita de modelos de adequabilidade de habitat. A diferença na extensão da área adequada predita por ambas resoluções foi medida de acordo a três limiares de corte diferentes. Os resultados indicaram diferenças na performance dependendo da resolução, com maior performance nos modelos de alta resolução. Além disso, a extensão de área adequada predita foi geralmente maior nos modelos de baixa resolução, com algumas exceções dependendo da espécie e do limiar de corte. Estes resultados ressaltam a importância da resolução das variáveis ambientais e suas implicações no uso de MAH no planejamento e manejo de Áreas Marinhas Protegidas (MPAs). / In this thesis I studied the deep-sea coral distribution along the Southwestern Atlantic Ocean and the effect of the environmental variables resolution on habitat suitability models (HSM). In the second chapter, I developed HSM based on the Maximum Entropy (MaxEnt) algorithm to identify the potential distribution of deep-water corals. To perform the models, I have used species presence data from existing datasets together with new species records and global environmental variables. I used the main environmental variables to identify suitable areas by modeling niches of different coral taxa. The studied taxa were Scleractinia and Octocorallia, as well as four reef-forming scleractinian species (Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata and Enallopsammia rostrata), and two non-reef-building scleractinians (Cladocora debilis and Deltocyathus spp.). Furthermore, I used multivariate approach to compare their environmental occupied niche in order to assess the spatial niche overlap and test niche similarity and equivalence. The resulting cold-water coral models showed that the most suitable areas occur along the west and southwestern continental margin of Brazil, which permitted to determine areas of priority for future exploration/research to determine conservation areas. Specific environmental niches suggested lower tolerance to high temperatures for Scleractinia than for Octocorallia. Besides, octocorals occupied a wider range of the environmental gradient, which was mainly related to particulate organic carbon, oxygen and temperature. This probably enables them to inhabit the northern part of Brazil where scleratinian models classified as poor habitats for this group. In addition, the Southwestern Brazilian continental margin encloses the most suitable habitats for deep reef-building species. This has important implications for conservation plans in that area already facing gas and oil exploitation and probably future mineral extraction. On the other hand, results suggest that different scleractinian cold-water coral species occupy similar environmental niches and, for this reason, I concluded that changes in environmental conditions, such as those associated to climate change, may affect them all. However, their niches are not equivalent and this suggests that each environmental factor could affect each species in different ways. In the third chapter, I performed HSM based on Maximum Entropy for Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata and Enallopsammia rostrata with two different environmental variables resolution, a low (i.e., a coarse resolution data with pixel size of ca. 1000 m2) and a high resolution (pixel of ca. 100 m2) to determine the effect of the environmental variables resolution on habitat suitability model performances and spatial extent of predicted suitable areas. From the final suitability maps, the changes in area extent depending of both resolution models were measured based in different thresholds. The results showed different performances of the models depending on data resolution, with higher performance at higher resolution. Furthermore, the predicted area varied between resolutions and threshold chosen. These results highlight the importance of environmental variables resolution and their implications for the use of HSM in planning and managing Marine Protected Areas (MPAs).

Na presente dissertação eu estudei a distribuição potencial dos corais de águas profundas ao longo da margem continental brasileira e o efeito da resolução das variáveis ambientais em modelos de adequabilidade de habitat (MAH). Primeiramente foi determinada a distribuição potencial de corais de águas profundas a partir de modelos de adequabilidade de habitat de máxima entropia (MaxEnt). Os modelos foram alimentados com diversas camadas de dados SIG de variáveis ambientais e dados de ocorrência de espécies extraídos de bancos de dados e de novos registros. Foram estudados os grupos Scleractinia e Octocorallia, assim como quatro espécies de corais formadores de recifes (Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata e Enallopsammia rostrata) e duas espécies de escleractínios não formadores de recifes (Cladocora debilis e Deltocyathus spp.). O nicho ambiental ocupado pelos táxons foi representado a partir de uma análise multivariada, sendo determinada a sobreposição entre os nichos. A partir disso foi avaliada a similaridade e equivalência de nicho ambiental entre as espécies. Os modelos resultantes indicaram que Scleractinia apresenta distribuição potencial na margem continental leste e sudeste do Brasil (porção Central e Sul), enquanto que Octocorallia apresenta uma distribuição potencial mais ampla abarcando toda a margem continental brasileira, com maior adequabilidade nas margens norte e nordeste. Assim, as áreas de distribuição potencial compartilhada entre octocorais e escleractínios foram aquelas ao largo da porção leste e sudeste do Brasil, sendo estas prioritárias para futuras pesquisas. Os nichos ambientais específicos indicaram menos tolerância de Scleractinia a altas temperaturas em comparação com Octocorallia. Os octocorais ocuparam uma maior amplitude do gradiente ambiental, principalmente relacionado às condições de carbono orgânico particulado, oxigênio dissolvido e temperatura. Isto possivelmente permite aos octocorais habitar a porção norte da margem continental brasileira, onde os modelos de escleractínios indicaram baixa adequabilidade de habitat. A margem continental sudeste do Brasil teve a maior adequabilidade para as espécies de escleractínios formadores de recifes profundos. Isto é de grande valor para planos de conservação nesta área que atualmente apresenta atividades de extração de óleo e gás e que, futuramente, pode ser explotada para a extração de minério. Além disso, as espécies formadoras de recifes ocuparam nichos ambientais similares e, portanto, mudanças nas condições ambientais, como por exemplo as associadas às mudanças climáticas, podem afetar todas elas. Contudo, seus nichos não são equivalentes e, portanto, cada uma delas faz um uso diferente do espaço ambiental da margem continental brasileira. Por último, foram desenvolvidos modelos MaxEnt para Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata e Enallopsammia rostrata em duas resoluções de análise, uma de baixa (i.e., com um tamanho de pixel de aproximadamente 1000 m2) e uma de alta resolução (i.e., pixel de aproximadamente 100 m2) para determinar o efeito da resolução das variáveis ambientais utilizadas na performance e na extensão de área predita de modelos de adequabilidade de habitat. A diferença na extensão da área adequada predita por ambas resoluções foi medida de acordo a três limiares de corte diferentes. Os resultados indicaram diferenças na performance dependendo da resolução, com maior performance nos modelos de alta resolução. Além disso, a extensão de área adequada predita foi geralmente maior nos modelos de baixa resolução, com algumas exceções dependendo da espécie e do limiar de corte. Estes resultados ressaltam a importância da resolução das variáveis ambientais e suas implicações no uso de MAH no planejamento e manejo de Áreas Marinhas Protegidas (MPAs). / In this thesis I studied the deep-sea coral distribution along the Southwestern Atlantic Ocean and the effect of the environmental variables resolution on habitat suitability models (HSM). In the second chapter, I developed HSM based on the Maximum Entropy (MaxEnt) algorithm to identify the potential distribution of deep-water corals. To perform the models, I have used species presence data from existing datasets together with new species records and global environmental variables. I used the main environmental variables to identify suitable areas by modeling niches of different coral taxa. The studied taxa were Scleractinia and Octocorallia, as well as four reef-forming scleractinian species (Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata and Enallopsammia rostrata), and two non-reef-building scleractinians (Cladocora debilis and Deltocyathus spp.). Furthermore, I used multivariate approach to compare their environmental occupied niche in order to assess the spatial niche overlap and test niche similarity and equivalence. The resulting cold-water coral models showed that the most suitable areas occur along the west and southwestern continental margin of Brazil, which permitted to determine areas of priority for future exploration/research to determine conservation areas. Specific environmental niches suggested lower tolerance to high temperatures for Scleractinia than for Octocorallia. Besides, octocorals occupied a wider range of the environmental gradient, which was mainly related to particulate organic carbon, oxygen and temperature. This probably enables them to inhabit the northern part of Brazil where scleratinian models classified as poor habitats for this group. In addition, the Southwestern Brazilian continental margin encloses the most suitable habitats for deep reef-building species. This has important implications for conservation plans in that area already facing gas and oil exploitation and probably future mineral extraction. On the other hand, results suggest that different scleractinian cold-water coral species occupy similar environmental niches and, for this reason, I concluded that changes in environmental conditions, such as those associated to climate change, may affect them all. However, their niches are not equivalent and this suggests that each environmental factor could affect each species in different ways. In the third chapter, I performed HSM based on Maximum Entropy for Solenosmilia variabilis, Lophelia pertusa, Madrepora oculata and Enallopsammia rostrata with two different environmental variables resolution, a low (i.e., a coarse resolution data with pixel size of ca. 1000 m2) and a high resolution (pixel of ca. 100 m2) to determine the effect of the environmental variables resolution on habitat suitability model performances and spatial extent of predicted suitable areas. From the final suitability maps, the changes in area extent depending of both resolution models were measured based in different thresholds. The results showed different performances of the models depending on data resolution, with higher performance at higher resolution. Furthermore, the predicted area varied between resolutions and threshold chosen. These results highlight the importance of environmental variables resolution and their implications for the use of HSM in planning and managing Marine Protected Areas (MPAs).

L’influence de la diversité génétique sur la stabilité démographique des populations constitue un paradigme de l’écologie évolutive. Au sein des populations naturelles, l’étude de cette relation est complexifiée par l’influence réciproque de la stabilité sur la diversité, et leur degré d’interconnexion. Ces interrelations ont été explorées chez la plante marine Zostera marina et les coraux d’eau froide Lophelia pertusa et Madrepora oculata, des espèces partiellement clonales. Ce trait d’histoire de vie, influençant profondément la dynamique démographique et la trajectoire évolutive des espèces, a constitué le fil d’Ariane de ce travail. L’échantillonnage dans l’espace (échelle régionale) et le temps (un pas de trois ans) d’herbiers de Zostère a permis de mieux comprendre la dynamique clonale de ces plantes. L’architecture et la diversité clonale apparaissent comme la résultante de l’équilibre entre dispersion/recrutement de nuages de graines dispersées collectivement, et la compétition pour l’espace entre clones. Les perturbations affectent localement l’équilibre de l’herbier. Cette dynamique originale rend impossible l’identification des contours populationnels. En revanche, nos résultats semblent indiquer que la diversité génétique au sens strict (hétérozygotie et nombre d’allèles) des herbiers de Zostères constitue un facteur de stabilité démographique, via sa potentielle influence sur les capacités de résistance aux perturbations saisonnières. Les coraux d’eau froide, quant à eux, présentent des patrons biogéographiques en accord avec l’hypothèse d’une extinction dans le Golfe de Gascogne, lors des derniers épisodes glaciaires. Les marques visibles des activités de pêche posent la question des capacités de résilience de ces écosystèmes, qui dépendent entre autres du potentiel de dispersion de ces espèces. L’absence de structure génétique observée chez L. pertusa suggère, au moins pour cette espèce, un fort degré d’interconnexion entre les récifs, tandis que M. oculata montre davantage de structure régionale. La sensibilité de ces espèces aux variations climatiques et à la pression des activités anthropiques souligne la nécessité d’études approfondies, pour leur conservation.Les résultats obtenus pendant cette thèse permettent de mieux comprendre la dynamique populationnelle des herbiers et récifs profonds, le taux de clonalité et la connectivité des populations. Ces informations sont essentielles pour avancer vers une meilleure compréhension de la dynamique et la résistance de ces espèces structurantes, et sont donc primordiales pour la conservation de ces écosystèmes clé. / The influence of genetic diversity on the demographic stability of populations constitutes a paradigm in evolutionary ecology. The complexity of this relationship within natural populations is enhanced by the reciprocal effect of stability on diversity, and the degree of interconnection among populations. This interaction was explored within the seagrass Zostera marina and the cold-water corals Lophelia pertusa and Madrepora oculata, three partially clonal species. This life history trait, deeply influencing the population dynamics and evolutionary trajectory of species, constituted the underlying theme of this work.The sampling in space (regional scale) and time (a three-years step) of eelgrass meadows allowed us to better understand the clonal dynamics of these plants. The clonal architecture and diversity may result from the equilibrium between dispersal/recruitment of collectively dispersed clouds of seeds, and the competition for space among clones. Perturbations locally affect the equilibrium of meadows. This original dynamic makes impossible the identification of population contours. Yet, our results suggest that the genetic diversity sensu stricto (heterozygosity and number of alleles) represents a factor of demographic stability, through its putative influence on resistance capacity for seasonal disturbances. Cold-water corals show biogeographic pattern in line with the hypothesis of glacial extinction, within the Bay of Biscay. The noticeable footprints of fishing activities question the capacity of resilience of these ecosystems, depending on dispersal potential of the structuring species, which showed low levels of clonality. The lack of genetic structure observed for L. pertusa suggest, at least for this species, a high degree of interconnection among reefs at large scale, while M. oculata revealed a stronger regional structure. Sensitivity of these two species to climatic variations and the pressures of human activities highlight the need of thorough studies for their conservation. Results obtained during this thesis allow a better understanding of the populations dynamics of both seagrass and deep reefs and their levels of clonality and connectivity. This information constitutes the first step toward a better understanding of dynamics and resistance of these structuring species, and is also primordial for the conservation of their key ecosystems.

One of the ocean’s greatest resources is the coral reefs, providing unique habitats for alarge variety of organisms. During drilling operations offshore many activities maypotentially harm these sensitive habitats. Det Norske Veritas (DNV) has developed arisk-based approach for planning of drilling operations called Coral Risk Assessment (CRA) to reduce the risk of negative effects upon cold water corals (Lophelia pertusa) on the Norwegian Continental Shelf (NCS). In order to get a good risk assessment amodelled dispersion plume of the drilling discharges is recommended. This study concerned a drilling case at the Pumbaa field (NOCS 6407/12-2) on the NCS, and used two different dispersion models, the DREAM model and the MUDFATE model in order to investigate how to perform good risk assessments. In the drill planning process a decision to move the discharge location 300 m north-west from the actual drilling location and reducing the amount of drilling discharges, was made inorder to reduce the risk for the coral targets in the area. The CRA analysis indicated that these decisions minimised the risk for the corals, and showed that the environmentalactions in the drill planning processes are necessary in order to reduce the risk for the coral targets and that the analysis method is a preferable tool to use. The amount of discharges, the ocean current data, the discharge location and the condition of the coral targets are the factors having the most important impact on the CRA results. From monitoring analysis from the case of study, it can be seen that a pile builds up around the discharge location. The dispersion models do not seem to take into account this build-up of a pile and thereby overestimate the dispersion of drilling discharges. This observation was done when modelled barite deposit was compared with barium concentrations measured in the sediment after the drilling operation. The overestimationis the case for the DREAM model, but has not been seen in the simulations with the MUDFATE model. Results from the modelling also indicated a higher overestimation for the DREAM model when using a cutting transport system (CTS) to release thedrilling discharges compared to release the discharges without using the CTS. / Korallrev består av ett skelett av kalciumkarbonat som bygger upp unika habitat på havsbotten. Dessa utnyttjas av flera olika organismer och är en av havets största och viktigaste resurser. Under prospekteringsborrningar till havs sker stora mängder utsläpp som kan påverka de känsliga miljöerna negativt. Det Norske Veritas (DNV) har utvecklat en riskbaserad strategi för planering av prospekteringsborrning i områden med koraller kallad Coral Risk Assessment (CRA). I CRA-analysen utvärderas risken för korallstrukturer (Lophelia pertusa) att påverkas av olika borrningsaktiviteter. Spridningsmodellering av det förväntade utsläppet från borrningsoperationen är ett viktigt hjälpmedel för att kunna utföra riskanalysen på ett tillfredsställande sätt. Studien har studerat en tidigare utförd prospekteringsborrning på Pumbaa-fältet (NOCS 6407/12-2) på den norska kontinentalsockeln och två olika spridningsmodeller DREAM och MUDFATE har jämförts i studien med syfte att förbättre riskbedömningen. I planeringsstadiet av prospekteringsborrningen togs ett beslut att flytta utsläppspunkten för det producerade borrslammet 300 m nordväst från brunnen samt att mängden borrslam skulle reduceras för att minska risken för påverkan på korallstrukturerna i området. CRA-analysen som utfördes i denna studie visade att dessa beslut minskat risken för korallstrukturerna att bli påverkade. Detta indikerar således att analysmetoden är ett viktigt verktyg att använda vid miljöundersökningar i planeringsstadiet för att minska risken för oönskad påverkan från aktiviteter i samband med prospekteringsborrning. De faktorer som har störst påverkan på CRA-analysen är mängden borrslam, strömdata, utsläppspunkt och tillståndet på korallstrukturerna. Under miljöövervakningen i samband med borrningsprocessen påvisades det att vallar av borrslam byggdes upp nära utsläppspunkten, vilket skedde relativt snabbt efter det att utsläppet startat. Spridningsmodellerna verkar inte ta hänsyn till denna uppbyggnad utan överestimerar spridningen och depositionen av borrslam. Detta har påvisats vid jämförelser av modellerade och uppmätta värden av bariumkoncentrationer i sedimentet. Överestimeringen är påvisad för DREAM, men slutsatsen är mer osäker för MUDFATE. Spridningsmodelleringen med DREAM indikerar även en större överestimering av resultaten om utsläppen sker med en så kallad CTS (Cutting Transport System).

L’acidification des océans provoquée par l’absorption du CO2 atmosphérique par l’eau de merest devenue une préoccupation écologique majeure et menace déjà les organismes calcifiants. Suiteà la révolution industrielle, le pH de l’océan de surface a diminué de 0,1 unité-pH. En revanche, celuides eaux profondes reste peu documenté. Les isotopes du bore (11B) dans les carbonatesbiogéniques se sont révélés être un puissant outil géochimique pour la reconstitution du pH, mais n’apas encore été appliqué aux coraux profonds. Un travail analytique sur MC-ICPMS Neptune et uneanalyse géochimique de ces coraux ont été effectués afin de reconstituer et quantifier l’acidificationdes océans. De même, la valeur 11B de l’eau de mer utilisée pour calculer les paléo-pH a étérevisitée et l’homogénéité des océans vérifiée.L’analyse de deux colonies modernes de Lophelia pertusa et Madrepora oculata a permis dequantifier un taux d’acidification pendant la fin du XXème siècle pour les eaux de sub-surface en mer deNorvège et cela après établissement d’une calibration expérimentale à partir de coraux de culture.L’analyse géochimique des coraux profonds a mis en évidence un effet vital lié à la physiologie quidoit être considéré pour quantifier avec précision la variabilité du pH. Celui-ci peut être en partiecorrigé par une analyse statistique des isotopes stables B, C et O. Cette étude a aussi révélél’influence de l’hydrodynamique régionale. Enfin les variations naturelles du pH pendant l’Holocène etle Dernier Maximum Glaciaire sur des coraux profonds fossiles méditerranéens ont été établies etcelles pendant l’aube de l’explosion de la diversité biologique. / Ocean acidification is caused by the absorption of rising atmospheric CO2 by seawater andrepresents a major environmental issue. Since the beginning of the industrial era, seawater pH hasdecreased by 0.1 pH units and is already threatening calcifying organisms. Boron isotopes (11B) haveproved to be a powerful geochemical tool for the reconstruction of pH variations, but has not yet beenapplied to deep-sea corals (DSC). Accurate and precise measurements of boron isotopes in coralsand seawaters were performed in order to measure small pH variations.The technique of pH reconstruction based on boron isotopes (pH-11B) was used on two specimens of the DSC Madrepora oculata and Lophelia pertusa collected alive in the Norwegian Sea and spanning an age of 40 (3) and 67 (3) years, respectively. Acidification rates were calculated by applying a new pH-11B calibration obtained from the geochemical analysis M. oculata and L. pertusa samples cultured under different pCO2 conditions. The contribution of a biological-related vital effect on d11B was observed at macrometer scale, and a correction was finally suggested based on oxygen and carbon isotopes. Overall, the coral δ 11B-based reconstructions show a pH decrease in the Norwegian Sea since the 1940s, which seems to be related to the local hydrodynamics. The pH-11B technique was also applied to fossil DSC fragments from two “on-mound sediment cores” retrieved in the Siculo-Tunisian Strait with the aim to reconstruct the pH during the Last Glacial Maximum and the Holocene periods. Finally, well-preserved limestone samples from the stratigraphic sequence Nama (551-543 Ma) in Namibia were investigated for 11B to study the pH variations at the beginning of the Cambrian evolutive radiation.

Graduation date: 2006 / Structure-forming invertebrates belong to a polyphyletic group of primarily sessile and sedentary megafauna that can significantly enhance the complexity of physical habitats. A number of these organisms, including cold-water corals and sponges, are known to be slow growing and vulnerable to physical disturbance. In addition, as filter feeders, these invertebrates can indicate areas of consistently favorable conditions for feeding and growth. This study provides the first quantitative analysis of structure-forming invertebrate communities in many areas along the continental margin of Oregon and Washington. Geological surveys during 1992-95, using the occupied submersible, Delta, sampled an extensive area in this region, primarily on and around rock outcrops. The videos from these surveys were analyzed to inventory and catalog sessile structure-forming invertebrates and to document their associations with geological habitat types. Detailed data on geological substrate, invertebrate diversity, abundance, and density were compiled and analyzed. It was found that geological substrate and depth were reliable indicators of suitable habitat for most species included in the study. Gorgonian corals tended to concentrate in high densities in depths between 200-250m, at the southern edges of submerged rocky banks, and where hard rocky substrate was covered with a thick layer of sediment. Because of recent fishery regulation changes, this information can be used as baseline data for future studies on the effectiveness of closed areas on the recovery of structure-forming invertebrates from disturbance, particularly bottom trawling.

Les habitats de coraux d’eau froide formés par des scléractiniaires coloniaux, des gorgones, des antipathaires et des pennatules sont des hotspots de biodiversité et de biomasse. Ils fournissent des fonctions importantes, comme des refuges et des zones d’alimentation, pour d’autres organismes. Mais, ils sont également vulnérables aux activités humaines, parce qu’ils sont fragiles, croissent lentement et atteignent des records de longévité. Dans les canyons sous-marins, le relief tourmenté, l’hydrodynamisme et l’hétérogénéité des substrats offrent des conditions environnementales propices au développement des habitats coralliens. Dans le Golfe de Gascogne, les coraux d’eau froide sont connus depuis la fin du 19e siècle, mais leur distribution, leur densité et leur rôle fonctionnel avaient été très peu étudiés.Pour augmenter cette connaissance, 24 canyons sous-marins et 3 sites sur l’interfluve/haut de pente contiguë aux canyons adjacents ont été visités par un ROV et une caméra tractée pendant 46 plongées au cours de 7 campagnes océanographiques. Les habitats coralliens définis par le système de classification CoralFISH ont été cartographiés à partir des images prises par les engins sous-marins puis la faune associée, les types de substrat et les déchets ont été annotés.Onze habitats coralliens et 62 morphotypes de coraux ont été observés dans les canyons sous-marins du Golfe de Gascogne hébergeant 191 morphotypes de faune associée, dont 160 morphotypes uniques. Les patrons de distribution à l’échelle locale et à l’échelle régionale pourraient être liés aux régimes hydrodynamiques et sédimentaires. Le type de substrat était important pour les assemblages de coraux et leurs faunes associées, distinguant les habitats biogéniques, sur substrat dur et sur substrat meuble. Les assemblages de coraux étaient similaires entre habitats biogéniques et habitats sur substrat dur, mais la faune associées était plus abondante et diversifiée sur les habitats biogéniques. La diversité-alpha, -beta et –gamma étaient étonnement élevée sur les habitats coralliens sur substrat meuble, égalant ou dépassant la diversité des habitats biogéniques.Les déchets marins étaient abondants et principalement composés de plastiques et de matériels de pêche. Ces déchets pourraient impacter les habitats coralliens : ils ont été trouvés à des profondeurs similaires et les déchets étaient piégés par des reliefs créés par des structures biologiques et géologiques. L’ocurrence des récifs de coraux préférentiellement sur les pentes plus abruptes des canyons sous-marins tandis que les débris de coraux sont plus fréquents sur des aires plus plates de l’interfluve ou du haut de pente, pourraient indiquer un impact de la pêche.Cette étude a contribué à l’initiative en cours de création d’un réseau Natura 2000 qui protégera à terme l’habitat « récif » dont les habitats coralliens biogéniques et sur substrat dur, mais pas les habitats coralliens sur substrat meuble. Pour ces derniers, un complément d’étude et d’autres stratégies de préservation seront nécessaires. / Cold-water coral (CWC) habitats formed by colonial scleractinians, gorgonians, antipatharians and sea pens are biodiversity and biomass hotspots that provide important functions, such as shelter and feeding grounds, to other organisms. But, they are also vulnerable to human activities, because they are long-lived, grow slowly and have a low resistance. Submarine canyons may offer the environmental conditions needed for CWC habitat development, due to their steep topography, complex hydrodynamics and substrate heterogeneity. In the Bay of Biscay, which margin is incised by hundreds of canyons, CWCs are known to exist since the late 19th century, but their distribution, density and functional role remained largely unknown, which impaired their preservation.To increase this knowledge, 24 canyons and three locations between adjacent canyons were visited with an ROV and a towed camera system during 46 dives on 7 cruises. Images were analysed for CWC habitats using the CoralFISH classification system. Within these habitats, corals, associated fauna were identified and substrate cover measured. Litter was identified in 15 out of 24 canyons.Eleven coral habitats constructed by 62 coral morphotypes were observed in the canyons of the Bay of Biscay hosting 191 associated megafaunal morphotypes, including 160 unique morphotypes. The distribution patterns at regional and local scales could be linked to hydrodynamics and sedimentary regimes. Substrate type was an important driver for coral and associated faunal assemblages, distinguishing biogenic, hard substrate and soft substrate habitats. Coral assemblages were similar between biogenic and hard substrate habitats, but the associated fauna was more abundant and diverse on biogenic habitats. The alpha, beta and gamma diversity was surprisingly high on soft substrate habitats, equalling or exceeding that of biogenic habitats.Marine litter was abundant and was mainly composed of plastic items and fishing gear. Litter could co-occur with CWCs and impact them: litter and most CWC habitats were observed at similar water depths and litter was more abundant in areas with a seafloor relief created by biological or geological features. Observations of coral reefs on steeper areas in the canyons and coral debris on flatter areas on the interfluve/upper slope may indicate a potential impact of the fishing industry. This study supports the ongoing effort to create a Natura 2000 network that will protect biogenic and hard substrate habitats, but also points out the need to develop a framework for the preservation of coral habitats on soft substrate.