Uncertainty is pervasive in all conservation decisions, the systematic treatment of which is necessary when evaluating the causes of species endangerment and deciding on appropriate actions to better ensure persistence. In Canada, a guiding principal of endangered species legislation for reducing uncertainty is adaptive management; however, adaptive management is not explicit in the development and application of most recovery strategies. To set the context for adaptive management recovery programming for at-risk species in Ontario, Canada, I synthesized existing knowledge, articulate and test critical uncertainties as hypotheses, and begin model development to predict expected outcomes of management alternatives, using the geo-politically defined population of Bobolink (Dolichonyx oryzivorus) in Ontario as a case study. Specifically, in Chapter 1, I combine evidence from empirical studies and expert judgement to model causal mechanisms driving Bobolink abundance dynamics and characterize the structural complexity of the management system. In Chapter 2, I spatially resolved Bobolink abundance trends to determine which areas of the province are contributing most substantially to overall abundance decline, and which presently satisfy recovery targets. In Chapter 3, I examined spatio-temporal variability in landscape pattern and processes to determine by which mechanism(s), and to what extent, changes in quantity and quality of agricultural grassland habitats have contributed to regional changes in abundance. Finally, in Chapter 4, I developed regionally scaled habitat suitability index (HSI) models to facilitate development of predictive population viability models. HSI models were used to evaluate to what extent lower-order proxies for processes of habitat selection influence patterns in abundance regionally. My research highlights how a science-based approach to recovery strategy planning can more effectively identify and address uncertainties in knowledge, and be used as an avenue to set the foundation of an adaptive management program.