Sprungmarken

 

Nitrogen and phosphorous cycling after four millenia of grazing

Felix Heitkamp1, Kessler Michael2, Sylvester Steven2, Jungkunst Hermann3
1 Landscape Ecology, Georg-August-Universität Göttingen
2 Institut für Systematische Botanik, Universität Zürich
3 Geoökologie/Physische Geographie, Universität Koblenz-Landau

O 1.3 in Long term trends in the functioning of ecosystems

14.07.2014, 11:55-12:15, H18

Human activities affected ecosystems worldwide to such an extent that there is even discussion that we have entered a new geological epoch, the Anthropocene. Although numerous studies quantified human impact on varies ecosystem processes and properties, the natural state of many ecosystems and their soils remains unknown. Moreover, human influence is manyfold across space and time and is therefore hard to quantify. In the high Andes (4500 m a.s.l.) paired plots were identified, one part under continuous grazing (rangeland) since at least four millenia, the other part pristine (bunch grass and forest) due to isolation by cliff-faces. In situ pedogenesis (commencing ca. 5 kyears BP) was indicated by gradually decreasing rock-fragments in the soil profile and fingerprinting with rare earth elements. Here we will show the integrated effects on nitrogen and phosphorous cycling of ecosystems developed with and without grazing pressure for at least four millenia. Stocks of C and N in the mineral soil were not significantly different in pristine and rangeland ecosystems. Stocks of P in rangeland soil (590 g m-2), however, increased  by a factor of 2 (bunchgrass, 270 g m-2) and 3.8 (forest, 160 g m-2). Pedogenetic mass balances revealed that pristine soils lost 40-90 kg P m-2 whereas rangeland soils accumulated 150 kg P m-2 during soil development. P-accumulation was attributed to P-import from remote areas. Microbial biomass C and N stocks were significantly higher in pristine soils (2 and 2.5-fold). Carbon-nutrient stoichiometry indicated P-limitation in pristine and N-limitation in rangeland soils. This indication was corroborated by much higher δ15N-values in rangeland as compared to pristine plants and soils. N-limitation in rangelands could be explained by hampered organic matter decomposition due to higher proportions of exchangable Al3+ and Al-humus associations. More than 50% of P were associated with amorphous pedogenic Fe/Al oxyhydroxides in rangelands and only one third in pristine ecosystems. Under the cold and semi-humid conditions of the sites, organic matter was apparently a more effective P-sink compared to pedogenic oxides. Analysis of more remote rangeland sites will reveal the long-term effects grazing under conditions of P-depletion and P-preservation (relative to pristine ecosystems) on nutrient cycling. With the current work, we could show for the first time differences in N and P-cycling in ecosystems, which developed from the beginning with and without grazing.



Export as iCal: Export iCal

last modified 2014-04-04