Understanding the effect of disturbance on rare and endangered species is critical for effective conservation. In this dissertation, I tested the effects of fire on the federally endangered Cape Sable seaside sparrow in Everglades National Park, Florida. In chapter 1, I used an unplanned fire event which burned a long-term research plot to test the effect of fire on the abundance and reproduction of a single subpopulation of the species. Sparrows were immediately impacted by the fire, in that no birds utilized the burned habitat for two full breeding seasons following the event. Moreover, no birds that had been banded in the burned habitat, prior to the fire, were ever relocated. Despite the initial impact, sparrows reclaimed burned habitat three years after fire at densities and nesting success indistinguishable from the unburned area. This suggests that fire provided neither benefit nor lasting negative effects to the sparrows. Vegetation structure was the most important factor in determining when sparrows return. In Chapter 2, I determined whether the processes witnessed at the local scale were supported in patterns of sparrow occupancy across the entire range of the species. I used a 16-year fire history database in conjunction with 13 years of survey data on sparrow site occupancy to calculate time-since-fire for each survey point. Then using logistic mixed models, I tested whether fire had an effect on occupancy and whether this effect varied according to time-since-fire. My results indicate that sparrow occupancy was significantly lower at points that had experienced fire one and two years prior, but this effect was lost in points with three years or more since being burned. My results, therefore, provide the first confirmation of fire effects on the Cape Sable seaside sparrow at both the local and landscape scales. In Chapter 3, I demonstrated how a long-term monitoring dataset, the 16-year sparrow helicopter survey, could be used to make inference about the statistical power of the current monitoring program. I used zero-inflated Poisson and binomial models to account for excess variation in the data, and generate parameter estimates from which I simulated sparrow population declines through time. Using these simulated data, I determined the statistical power of the current sparrow survey to detect meaningful declines in both abundance and occupancy. My results showed that the current sparrow survey is unable to detect even large declines (>90%) over short time periods (three years). Survey power increased with additional years of data (5-10 years). With efforts to restore the Everglades ecosystem currently underway, park managers require a more precise tool than the current survey in order to detect important changes to sparrow populations.