In 2011 and 2012, I had the privilege of visiting the Xinjiang Institute of Ecology and Geography in the far north-west of China, and I’ve been collaborating with this research group ever since. The region’s capital city Urumqi is one of the most continental places on earth – it is the city the furthest away from any ocean (> 7000 km). The landscapes here are unlike anything that I had ever seen in the Pacific or in Europe. Large parts of the province of Xinjiang are covered in deserts, including the Taklamatan, Gobi and Gurbantünggüt deserts.
Arid and semiarid ecosystems comprise more than 40% of the earth’s terrestrial surface, and these systems sequester a lot of carbon; they are a lot more productive than they look, contributing 30-35% of the global net productivity. In fact, recent evidence suggests that the variation of the global carbon cycle from year-to-year will be less and less due to tropical rainforests and more due to semiarid ecosystems. However, remarkably little is known about the things that (may) affect the carbon cycle of these systems.
One obvious driver of the composition and productivity of desert ecosystems is precipitation, because only after rain, the vegetation appears to ‘become alive’. In central Asia, desert vegetation communities can typically be divided into shallow-rooted and deep-rooted species, which have different water-use strategies. Previous work of my colleagues showed that the main component of the community (a shrub named Tamarix) is one of these deep-rooted species that relies on groundwater throughout the whole growing season, and that its productivity doesn’t respond to precipitation. So if it isn’t the shrubs that are responsible for the huge variation in productivity in response to rainfall, what is?
A previous study I was involved with showed that years with increased precipitation before the growing season were associated with an increased cover of short-lived shallow-rooted plants. So we wanted to find out if & how this change in community structure could lead to huge differences in the whole-year carbon budget – and we used a combination of a very long-term gas exchange dataset and several experiments to do this. The results of this study were recently published in:
Liu R, Cieraad E, Li Y, Ma J 2016. Precipitation pattern determines the inter-annual variation of herbaceous layer and carbon fluxes in a shrub-dominated desert ecosystem. Ecosystems 19 (In Press).
Both the long-term observations and experiments showed that sufficient pre-growing season precipitation (between November to April, mostly falling as snow) led to more germination and growth of shallow-rooted annual plants. Only if this was followed by high-frequency rain during the growing season, were these plants able to survive and use that rain to increase community productivity. This would lead to community productivity as high as has been recorded in grassland and forest ecosystems.
So, it is the little guys, like grasses and these annuals below, that make the difference in turning this desert from a source into a sink of carbon – but only if it snows and rains enough, AND at the right time!
This study shows how important year-to-year variation in precipitation and community composition is to the amount of carbon stored or released. It also highlights that any predicted changes in the frequency and seasonal pattern of precipitation in this region (and elsewhere) due to global change will have important ramifications for the regional and global carbon cycles.