PS 50-57
Does plant biodiversity mediate land management impacts on grassland soil carbon?

Thursday, August 14, 2014
Exhibit Hall, Sacramento Convention Center
Pete Manning, Institute of Plant Sciences, University of Bern, Bern, Switzerland
Ingo Schöning, Max Plank Institute for Biogeochemistry Jena
Marion Schrumpf, Max Plank Institute for Biogeochemistry Jena
Steffen Boch, Institute of Plant Sciences, University of Bern, Bern, Switzerland
Eric Allan, Institute of Plant Sciences, University of Bern, Bern, Switzerland
Wolfgang W. Weisser, Chair of Terrestrial Ecology, Technical University of Munich, Freising, Germany
Markus Fischer, Institute of Plant Sciences, University of Bern, Bern, Switzerland
The Biodiversity Exploratories Consortium, Senckenberg Gesellschaft für Naturforschung, BIK-F - Biodiversity Exploratories Office, Frankfurt, Germany
Background/Question/Methods

While it is known that soil carbon can be affected by a range of factors, including climate, land management, plant community composition and abiotic soil properties, the relative contribution of these to landscape variation in soil carbon stocks is poorly understood. As many of the drivers of soil carbon stocks co-vary, or are typically studied in isolation, it is difficult to disentangle their effects and estimate their relative importance. Thus, it is unclear whether the positive relationship between plant biodiversity and soil carbon storage, as reported from controlled experiments, is an important mechanism in real-world ecosystems.

We utilized structural equation modelling to understand the relationships between and relative importance of numerous drivers in determining landscape variation in soil grassland carbon stocks. This was achieved using data from a large-scale and long-term research platform, the German Biodiversity Exploratories. In 1500 grassland plots from 3 regions of Germany soil carbon stocks were measured along with numerous management factors and abiotic soil properties. Plant community data from these same plots was also used in conjunction with plant trait values from databases to generate a range of measures describing functional properties of the vegetation. 

Results/Conclusions

We show that soil carbon in the top soil layers is strongly influenced by land management and that these effects operate both directly and via changes to the functional composition of the vegetation. In deeper soil layers the effects of management and plant community properties are less pronounced or absent and abiotic soil properties play a relatively greater role in explaining soil carbon stocks. These results suggest that land management aiming to boost soil carbon may have little effect upon deep carbon. They also demonstrate that management of plant community composition has the potential to boost soil carbon stocks, both in in its own right and by modifying the effects of other management factors.