Livestock productivity as indicator of vulnerability to climate hazards: a Mongolian case study

Mongolia is subject to regular peaks of livestock winter mortality called dzuds. Several kinds of dzud exist and the 'white dzud', characterized by heavy stochastic snowfalls preventing livestock to access forage, is considered the most common. Droughts and high livestock densities are thought to be part of the dzud process by affecting body condition, which increases vulnerability to snowfalls. Guided by the equilibrium/nonequilibrium framework, we studied how rainfall, animal numbers and pasture health (defined as the integrity of ecological processes sustaining grass growth) impact livestock body condition in a case study of West Mongolia. We studied this parameter through livestock productivity (LP) as a proxy, defined as the annual number of newborns per breeding-age female. We found no significant impact of rainfall or livestock numbers, alone or combined. We found through the study of pasture use, defined as the ratio forage consumed/forage available, an impact of the combined effect of rainfall, animal numbers and pasture health. We observed in addition sharp LP decreases prior to dzuds, which suggests that the above-mentioned drivers interact to weaken livestock which increases its vulnerability to winter hazards. This tends to show that in our case study, dzuds are not the simple consequence of stochastic hazards striking randomly, but instead, the final stage of a chain of events that involves dry years, high livestock densities and pasture degradation. This also indicates that dzud early warning indicators could be designed based on LP monitoring.