Fisheries managers have the potential to significantly improve reef fish management in the Gulf of Mexico through the use of ecosystem-based approaches to fisheries management. Ecosystem-based approaches are needed to address the effects of fishing on trophodynamic interactions, to better account for ecosystem-scale processes in model projections, and to recognize the short and long-term biomass tradeoffs associated with making regulatory choices. My research was concentrated around three objectives: (1) characterizing the trophodynamic interactions between Gulf of Mexico fishes, in order to construct an invaluable tool (a Gulf of Mexico Atlantis model) to be used in ecological hypothesis testing and policy performance evaluation for years to come; (2) predicting ecological indicators for the Gulf of Mexico that both respond to fishing pressure and are robust to observational error, and; (3) evaluating the performance of an ecosystem-based policy options for managing reef fish species in the Gulf of Mexico. To accomplish these objectives, a spatial, trophodynamic ecosystem model- Atlantis, was employed to represent the Gulf of Mexico marine ecosystem. To characterize trophic interactions between modeled species, I applied a maximum likelihood estimation procedure to produce Dirichlet probability distributions representing the likely contribution of prey species to predators’ diets. This provided mode values (the peak of the distribution) and associated error ranges, which describe the likely contribution of a prey item in a predator’s diet. The mode values were used to parameterize the availabilities (diet) matrix of the Gulf of Mexico Atlantis model. Investigating trophic interactions was useful for determining which species within the Atlantis model were data rich, and justified the emphasis on reef fish species and their prey items in subsequent analyses. Once parameterized and calibrated, I used the Atlantis model to project ecological indicators over a 50 year time horizon (2010-2060) under varying levels of fishing mortality. Principal component analysis was used to evaluate ecological indicator trajectories in multivariate space, to rank indicators according to how well they describe variability in ecosystem structure (termed ‘importance’), to reveal redundancies in the information conveyed, to quantify interannual noise and to determine how robust indicators are to observational error. Reef fish catch, Red snapper biomass, King mackerel biomass and Species richness indicators ranked the highest in terms of importance and robustness to error and in having low levels of interannual noise (i.e., requiring less frequent monitoring). I then used a management strategy evaluation (MSE) framework in Atlantis to evaluate some of these same indicators under an ecosystem-based approach to fisheries management – using robust harvest control rules to manage reef fishes. I found that this ecosystem-based policy option was able to maintain higher reef fish biomass, catch and ecosystem-wide biodiversity under any given level of fishing mortality when compared to a status quo management approach. These results suggest that harvesting under the HCRs encourages an alternative ecosystem state with a more Pareto-efficient tradeoff frontier than the status-quo policy. A potentially reduced extinction risk for reef fish is plausible under this ecosystem-based policy option. This research provides a quantitative look at the fishery performance and ecological tradeoffs associated with various policy options. MSE methodology using ecosystem-based policy performance metrics is also demonstrated. Tool development and findings from this research should aid in the development of ecosystem-based policies for this region.

Submitted by Rafael Santana (rafael.silvasantana@ufpe.br) on 2017-07-14T18:46:22Z No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Dissertação_Walter DM Oliveira_ Oceanografia_UFPE.pdf: 1745886 bytes, checksum: 62870c6bfbb9848afc84e4fb94eff45e (MD5) / Made available in DSpace on 2017-07-14T18:46:22Z (GMT). No. of bitstreams: 2 license_rdf: 811 bytes, checksum: e39d27027a6cc9cb039ad269a5db8e34 (MD5) Dissertação_Walter DM Oliveira_ Oceanografia_UFPE.pdf: 1745886 bytes, checksum: 62870c6bfbb9848afc84e4fb94eff45e (MD5) Previous issue date: 2016-05-31 / Os recifes costeiros são comumente afetados por ações antrópicas que vão desde o turismo contemplativo a poluição por efluentes domésticos. Essas formações recifais abrigam muitas espécies de peixe que em busca de abrigo para diferentes fazes do ciclo de vida e alimento. Nesses sentido as cristas recifais formam um importante papel para o desenvolvimento desses organismos, sem comumente um local de berçário natural. A cobertura de substrato é um outro fator importante para essas espécies pois pode gerar proteção, alimentação direta a indireta com invertebrados epifíticos. As cristas recifais do litoral sul de Pernambuco são dominadas por cobertura algal. Os recifes locais são importantes como fonte de renda para as comunidades locais e para conservação dos ambientes marinhos costeiros. Foram analisadas as relações entre a ictiofauna e a categorias de algas nas cristas recifais de Porto de Galinhas, Serrambi e São José da Coroa Grande, tendo sido amostrados três piscinas em cada praia. Para isso foram feitas amostragens com transecto de faixa de 20x2 m² em cada piscina, com registros fotográficos da cobertura de substrato, essas fotos foram posteriormente analisadas para estimar o percentual de cada categoria representada na cobertura do substrato da crista recifal. Os dados mostraram que Stegastes fuscus, Abudefduf saxatilis, Sparisoma axillare, Halichoeres poeyi e Ophioblennius trinitatis foram as espécies mais representativas nas cristas recifais analisadas. As algas calcárias articuladas e turf foram as algas mais abundantes nessas formações. As algas folhosas foram pouco abundantes, sendo mais representadas em São José da Coroa Grande. A maioria da abundância de peixes representou a categoria trófica de herbívoros errante e a mais rica foi comedores de invertebrados móveis. St. fuscus e O. trinitatis foram registrados em todas as classes de tamanho, enquanto A. saxatilis, Sp. axillare,e H. poeyi foram observados em fases juvenis. A relação verificada entre a comunidade de peixes, a cobertura de substrato e as características estruturais foi baixa visto que a distribuição de peixes foi homogênea e houve pouca variação nas categorias de algas. / The reef crests form an important habitat for fish of different habits and life stages. Coastal reefs are commonly affected by human activities ranging from the contemplative tourism to the pollution from domestic sewage. The substrate coverage can indicate the distribution of fish species that use this place to establish throughout the life cycle or in transitional phases. The reef crests of the southern coast of Pernambuco are dominated by algal cover. Local reefs are important as source of income for local communities and for conservation of coastal marine environments. The relationship between fish community and categories of algae on the reef crests of Porto de Galinhas, Serrambi and São José da Coroa Grande were analyzed. Three points were sampled at each beach. For the sampling of fish were made four transects of 20x2 m in each pool. At the same time were made photos of substrate coverage, these photos were then analyzed to estimate the percentage of each category. The data showed that Stegastes fuscus, Abudefduf saxatilis, Sparisoma axillare, Halichoeres poeyi and Ophioblennius trinitatis were the most significant species in the reef crests analyzed. The jointed calcareous macroalgae and turf were the most abundant in these formations. The sheet-like algaes were less abundant and more represented in São José da Coroa Grande. Most of the abundance of fish represented the trophic category herbivores and the richest was eaters of mobile invertebrates. St. fuscus and O. trinitatis were recorded in all size classes, while A. saxatilis, Sp. Axillare, and H. poeyi were observed in juvenile stages. The relationship between the fish community and substrate coverage and structural characteristics was low as the fish distribution is homogeneous and there was little variation in types of substrate coverage.

The objective of this project was to determine if the electromagnetic field (EMF) emissions from undersea power cables impacted the local and transient marine life, with an emphasis on reef fishes. The work was done at South Florida Ocean Measurement Facility of Naval Surface Warfare Center, Carderock Division, Broward County, Florida. This facility functions as the hub for a range of active undersea detection and data transmission cables. It has multiple active submarine power cables that extend several miles offshore and which can deliver power and enable data transmission to and from a range of acoustic and EMF sensors. The cables lie directly on the seabed, are buried in the sand, or are suspended in the water column. EMF emissions from a selected cable were created during SCUBA fish surveys. During the surveys the transmission of either alternating current (AC) or Direct Current (DC) was randomly intiated by the facility with no transmitted current (OFF) provided a control. The surveys were conducted using standardized transect and stationary point count methods to acquire reef fish abundances prior to and immediately after a change in transmission frequency (the divers were aware of the time of frequency change but not the specific frequencies). The divers were also tasked to note the reaction of the reef fishes to the immediate change in the EMFs emitting from the cable during a power switch. The surveys were conducted on a quarterly basis at three sampling sites offshore on the same cable. These sites were in water depths of approximately 5, 10, and 15 m, respectively and were selected based on their robust reef fish community and are representative of each of the three primary hardbottom coral reef habitats in the local offshore environment: the Inner (Shallow), Middle, and Outer (Deep) reef tracts. A total of 263 surveys were conducted: 132 transect-counts and 131 point-counts over 15 months. There were 24,473 fishes counted during transect-count surveys and with point-counts, 36,115 fishes were counted. With count types and sites combine a total of 151 species representing 35 families were recorded. An analysis of the data primarily did not find statistical differences among power states and any variables. However, this may be a Type II error as there are strong indications of a potential difference of a higher abundance of reef fishes at the sites when the power was off. There are a number of caveats to consider with this finding: the data set needs to be larger in terms of numbers of: counts, sites and eletro-sensitive species to allow for rigorous statistical analysis; also a longer time between frequency changes to allow for slower, but nonetheless important, reactions to differing EMFs might lead to differing conclusions. Obviously, more research is required to confirm the results of this study.