Remote sensing of ecosystem vulnerability: Assessing climate-vegetation-livestock interactions in Mongolia

Abstract:

Stock breeding is a major economic sector of Mongolia, supporting unique cultural and social identity. In spite of its long history, contemporary pastoralism increases interventions on climate-vegetation interactions substantially, which results in negative feedbacks to livestock sector. This presentation draws an attention how natural processes of climate and vegetation interact with livestock dynamics. Massive loss of livestock and wildlife animal during winter seasons (dzud) is an endemic climatic disaster in the Central Asia grasslands but the mechanisms are not well understood yet. Recent national-wide sever Dzud occurred during 2009-2010 winter in Mongolia. The dzud mechanisms were investigated by developing a schematic mechanism model on climate-vegetation-livestock interactions and applying it for quantitative statistical analysis. Various remote sensing products were integrated to prepare the status and process variables of the schematic model, including daily temperature, precipitation, evapotranspiration, and primary production and biomass for a period from 2003 to 2010. At a lower level of administration (i.e., ‘soum’ generally larger than 1000 km²), stepwise multiple regression analysis was conducted to find significant factors of inter-annual livestock change. As results, linear regression models were successfully produced at 70% of soums. Summer and winter variables appeared equally important in controlling livestock dynamics. The primary factor of each soum showed certain regional patterns incident well with climate severity and foraging resource availability (e.g. temperature in north, dryness in south, and NDVI in middle). Regional pattern of herbaceous biodiversity depends on both climate and disturbance (i.e. fire and grazing) gradients but the livestock grazing effect appeared localized normally within 1.5 km from livestock shelter or wells. At a local-scale (i.e. family level smaller than 100 km²), species composition seems to provide useful indicator of grazing pressure, while climate and fire disturbance determined regional pattern of vegetation biodiversity. The results provide a useful premise to devise a satellite-based assessment tool for foraging resource availability and biological regime shift by grazing and climate change in future study.