| Human land use is widely acknowledged to be the most influential disturbance agent in ecosystems around the globe. Changes in the disturbance regime profoundly affect components of natural and coupled socio-ecological systems such as land cover, ecosystem processes, and biodiversity. Excessive increases in disturbance may lead to catastrophic shifts in the composition and function of communities, as critical thresholds of disturbance are reached. Decreases in land use intensity are less common, and less studied, but no less important and provide opportunity to understand ecosystem recovery.;The overarching goal of my dissertation was to study the effects of decreasing land use intensity on fire, vegetation, and wildlife dynamics in arid ecosystems. I chose the south of European Russia as my study area because the collapse of the Soviet Union here led to rapid changes in the main land use -- sheep grazing -- which caused substantial environmental changes. I used a variety of remote sensing data and image processing methods to quantify scale and pattern of these changes from 1985 until 2007 and to link these changes with habitat selection of Saiga antelope --- a species of major conservation concern. Results revealed significant changes in ecosystem patterns and processes following the collapse of the USSR. Decrease in land use intensity led to dramatic increase in fires from almost none in the 1980s to more than 20% of the total study area burned annually in the 2000s. It took vegetation 5 to 6 years to recover the point where large scale fires could be sustained. The total area burned was driven by a number of factors, including sheep numbers, precipitation during growing season, temperatures during burning season and vegetation condition. Together with fires, vegetation composition changed substantially, switching from communities dominated by shrubs to perennial grasses. Though current habitat selection by Saiga antelope was not affected by vegetation composition at the scale of my analysis, fires played prominent role as a landscaping agent and areas that burned more often and were close to water sources were preferentially selected. In general, my dissertation shows that institutional changes may lead to both ecosystem recovery and transitions to unexpected states. Natural experiments set up by large scale socio-economic changes can provide a framework to study the effects of such changes over large areas and elucidate the direction of ecosystem development under decreasing land use conditions in other parts of the world. |
