Tracking recovery in disturbed environments requires the consideration of many spatial and temporal scales. A sensitive indicator organism, Hyalella azteca, was used to assess lake recovery at multiple scales in the region of Sudbury, Canada. A 40-lake presence-absence survey was conducted over a period of 23 years to track colonization history and chemical factors that limit this typically ubiquitous organism. A six-lake study was used to investigate the importance of spatially varying watershed characteristics within lakes, which could provide habitat hot spots for colonization during early stages of recovery. An intensive single-lake study examined the effects of local-scale chemistry (e.g., bioavailable metals, waterborne organic matter) and adjacent subcatchment terrestrial features on the availability of suitable habitat for colonizing amphipods. At the regional scale, presence-absence models suggested that colonization probability increased with lake water conductivity and alkalinity. Within lakes, subcatchment confluence sites appeared to be important habitats in the early stages of colonization. Site-specific features, such as macrophyte and woody debris cover, increased and decreased H. azteca abundance, respectively, and yet these relationships were influenced by adjacent terrestrial subcatchment characteristics. For example, with more terrestrial vegetation, the relative increase in abundance due to macrophyte cover was further increased. Within the intensively-studied lake, larger subcatchments with more terrestrially-derived waterborne organic matter had higher abundances of H. azteca. Using H. azteca as an indicator of aquatic ecosystem health, these relationships suggest that as lakes recover, subcatchment confluence sites can be hot-spots for colonization, and their suitability improves with interactions between local habitat characteristics and terrestrial characteristics. / Thesis (Master, Biology) -- Queen's University, 2014-01-31 21:50:26.021

Os sistemas naturais e humanos são considerados sistemas integrados, com interações complexas e de caráter fortemente multidisciplinar. A abordagem da limnologia da paisagem, como um princípio holístico para avaliar as relações complexas entre a bacia de captação e o reservatório, é de grande importância na produção de diagnósticos consistentes. O modelo de tomada de decisão foi produzido por meio de literatura especializada, conhecimento de especialistas, Processo Hierárquico Analítico (AHP) e Avaliação Multicritério (MCE). O modelo foi desenvolvido para atuar em escala da paisagem, considerando a bacia de captação como a escala observacional. Foram selecionados sete critérios (variáveis preditoras) para determinar o grau de fragilidade ambiental para o processo de eutrofização antrópica em reservatórios tropicais: Variáveis Intrínsecas do Reservatório: (1) morfometria (profundidade) e Tempo de Retenção (RT); (2) Variáveis Limnológicas: Zona Eufótica e presença de macrófitas de crescimento intensivo; (3) Variáveis Antrópicas: Fonte de Poluição Difusa Potencial; e (4) Variáveis Hidrológicas: Taxa de Sedimentação e escoamento superficial potencial. No processo de ponderação das variáveis, todas as matrizes foram consistentes e os especialistas priorizaram o Escoamento Superficial potencial e a Fonte de Poluição Difusa Potencial com os principais responsáveis pelo processo de eutrofização antrópica. Apesar dos critérios e pesos serem fixos para os reservatórios tropicais, existe a possibilidade de ajuste do modelo de acordo com situações especiais, já que o modelo é flexível suficiente para ser utilizados em outras bacias de captação, com características diferentes e intrínsecas. O modelo também é adaptável em função da disponibilidade de base de dados / The human and natural systems are integrated systems, with complex interactions and a strong multidisciplinary character. The application of landscape limnology, as a holistic principle to address the complex relationships between the watershed and reservoir, are of great importance to produce consistent diagnoses and predictions. The GIS-based model was performed by using literature, expert knowledge, Analytical Hierarchy Process (AHP) and Multi-Criteria Evaluation (MCE). The model was performed in a landscape scale considering the watershed as the observational scale. We selected seven criteria (predictor variables) to assess the environmental fragility to human-induced eutrophication in tropical reservoirs: (1) Intrinsic Reservoir Variables: morphometry (depth) and Retention Time (RT); (2) Limnological Variables: euphotic zone and intensive-growth macrophytes presence; (3) Anthropic Variable: Potential Non-Point Source (NPS); and (4) Hydrological Variable: Sedimentation Rate and Potential Runoff. In the weighting process, all matrices were consistent and the experts prioritized the potential runoff and potential NPS criteria as the main drivers of human-induced eutrophication. Although the criteria and its weights are considered fixed for any tropical reservoir, it is possible to adjust them according to specific situations since the model proposed is flexible enough to be used in different watersheds with different and intrinsic characteristics and adapted according to the database availability