Drivers of phosphorus cycling in temperate and boreal forest soils

Understanding the main underlying drivers of soil nutrient cycles is essential for predicting the effects of global change on them. The overall aim of this thesis was to gain insights into the mechanisms of phosphorus (P) cycling in unmanaged forest ecosystems and to study how these mechanisms are linked to soil formation and global environmental changes. The soil P cycle was studied by quantifying P pools and fluxes. The influence of several factors of soil formation on P cycling was assessed at different experimental scales, ranging from a laboratory incubation, via semi-controlled field experiments, to studies at the ecosystem scale. First, litter decomposition rates were studied to assess the magnitude and controlling factors on nutrient release and microbial functioning. Litter of differing carbon:nitrogen:phosphorus (C:N:P) ratios was used and incubated in a laboratory study at 5°C and 15°C. Litter samples were taken along a natural gradient of plant growth forms in two forest-tundra ecotones in northern Russia. This study revealed that microorganisms, major mediators of litter decomposition, were able to adapt to a wide range of nutrient conditions by adjusting the C:N:P ratios of their biomass. The carbon use efficiency (CUE) of microorganisms decreased with increasing litter C:N:P ratios from forest to tundra, indicating the occurrence of “overflow respiration” where microorganisms respire C to acquire nutrients. Net N and P release was negatively correlated with litter C:N:P ratios, with almost no release of N and P from tundra litter. Experimental warming by 10°C had little influence on the nutrient release, indicating that litter quality is more important for nutrient release than temperature. The observed increase in nutrient release from tundra to forest strongly suggests that nutrient release contributes to the stability of the present vegetation types. The lower nutrient release in the tundra than in the forest potentially reduces plant growth and thereby decreases the forests’ ability to expand into tundra. Second, a field study on P cycling in two German beech forests demonstrated for the first time significant phosphatase activity in percolating soil solution from forest soils. As a consequence, there is a subsequent quick hydrolysis of phosphate-monoesters to phosphate in sampled water, which prevents a clear determination of the ratio of monoesters to phosphate present within the soil. Based on these results, it was shown that phosphatase activity leads to an underestimation of labile P in soil solution, with concentrations more than halving within 24 hours of sample storage. The overestimation of phosphate mobility in the soil, and the conclusion that soil P depletion is most likely driven by non-labile dissolved organic P leaching is highly relevant for accurately predicting long-term P processes. Third, the influence of soil nutrient status, fertilization, and seasons on the leaching of inorganic and organic P and N was studied over 15 months in the field. For this, two beech forest soils differing in total P stocks and soil type were examined. The study revealed that the effects of N and P fertilization on N and P leaching strongly depended on the N and P status of the soil. Fertilization increased the leaching stronger at the low-P site. Also, drying and rewetting increased P leaching more strongly at the low-P site. This suggests that the nutrient-poor soil is more susceptible to environmental changes like in the temperature and precipitation regime or in the atmospheric composition. Further, P forms were differently affected by the N and P fertilization and seasonal changes, with dissolved inorganic P leaching being particularly sensitive to changes in weather conditions, e.g. rewetting after longer dry spells. Comparisons of input and output fluxes of P revealed that at these two Abstract 6 sites under the prevailing conditions, leaching losses from the mineral topsoil were comparable in magnitude to atmospheric P inputs. Finally, the impact of plant-soil interactions on the P cycling was examined in an unmanaged tundra-forest ecosystem on the Kola Peninsula, Russia. This study showed that indirect temperature effects resulting in differing vegetation types, had a stronger impact on N and P cycling than direct temperature itself. Despite similar total N and P stocks and the same age of soils, the distribution of P pools differed across the forest-tundra ecotone. Available nutrient stocks increased from tundra to forest, despite a strong concomitant increase in plant biomass, while the P pool in stones in the soil profile decreased, hinting to an increased weathering in the forest. Moreover, the available nutrient stocks (ammonium and resin extractable P) were also higher in soils under tree canopies than in open patches of the forest, implying that higher available nutrient stocks were related to the presence of trees and not only due to climate. Potential mechanisms include enhanced weathering in tree rhizospheres. Further, higher nutrient availability resulted in higher concentrations of leaf nutrients and hence, increased nutrient release during decomposition, potentially reinforcing the observed vegetation patterns. Soil microorganisms adapted to the gradient in available nutrients along the transect by adjusting their C-to-nutrient ratios and producing more extracellular enzymes in the nutrient-poor tundra compared to the boreal forest. The interplay of biotic and abiotic processes resulted in faster soil development in the boreal forest compared to the tundra, indicating the possibility for a positive feedback of treeline advance with warmer climate. In conclusion, this thesis showed that P cycling in unmanaged forest ecosystems is strongly shaped by its nutrient status, driving long-term soil development. Phosphorus cycling is significantly altered and, most likely, accelerated by the combined impact of global change on the frequency of extreme weather events and on inputs of C, N, and P. Nitrogen and P inputs were shown to be important drivers for the P cycling, but the magnitude of the effect depended on the current nutrient status of the soil and the chemical nature of P. This applies, in particular, to organic P, as leaching of non-labile dissolved organic P was identified as a potential main driver of long-term soil P depletion. Phosphorus cycling is also strongly affected by vegetation, as biogeochemical processes in the rhizosphere depend on functional plant types, leading to positive and negative feedbacks. This is supported by the finding that litter quality, being a result of plant-soil interactions, is more important for nutrient release than temperature itself. Also, this thesis gives evidence that processes in nutrient poor soil-plant systems are more susceptible to environmental changes such as extreme weather events or the increase of C, N, and P inputs. In summary, it is crucial to explicitly consider the nutrient status of soils in studies and models investigating soil processes, and to focus on organic and inorganic P, as our understanding of the P cycle still lags behind our understanding of the one of C and N cycles