Interactions between plants and soil nutrient cycling under elevated CO2

The atmospheric concentration of the greenhouse gas CO2 is rising and may stimulate plant production and soil C input. If soil C input rates exceed soil C respiration rates under elevated CO2, global warming may be mitigated by long-term soil C sequestration. However, whether soils will serve as CO2 sinks is still debated, since it is uncertain how elevated CO2 will affect the interactions between plant growth and soil nutrient cycling. In the first part of this dissertation, I explored how long-term elevated CO2 affects soil C inputs versus SOM decomposition, and how these changes ultimately feedback to soil C sequestration. This research was carried out in a Free Air Carbon dioxide Experiment (FACE) in Switzerland that had been exposed to elevated CO2 and N fertilization treatments for 10 years. The isotopic label of the applied CO2 and N allowed for tracing new C and N dynamics in the system. In addition, I summarized available data related to plant growth and soil nutrient cycling from long-term CO2-enrichment experiments using the statistical tool Meta analysis. By incubating litter and soil derived from Swiss FACE, I concluded that the impact of elevated CO2 on litter quality and litter decomposition rates was minor. Therefore, elevated CO2 is not expected to affect soil C contents through its impact on litter quality and decomposition. The Meta analysis showed that the main driver of soil C sequestration is not SOC decomposition, but soil C input through plant growth, which is strongly controlled by nutrient availability. If soil nutrient availability was high, soil C input outweighed C decomposition leading to net C sequestration. However, if soil nutrient availability was low, soil C input rates lagged behind soil C decomposition rates due to CO2-induced nutrient immobilization, which had reduced plant growth. Thus, for soil C sequestration under elevated CO2 ample soil nutrient availability is required. In the Swiss FACE experiment however, soil C sequestration did not increase under elevated CO2, despite high fertilization rates, concurrent increases in plant growth, and relatively low decomposition rates. This may be due to frequent harvests and shows that the potential for soil C sequestration in individual agro-ecosystems is still uncertain, due to management practices that can affect soil C input and/ or soil C decomposition. The potential for soil C sequestration in individual unfertilized/ natural ecosystems is also unclear, since unexplained processes appear to prevent N limitation in some of these FACE systems. These processes may occur in the rhizosphere, which is often overlooked, but plays a vital role in mechanistically coupling plant production and soil nutrient cycling. In the second part of this dissertation I focused on how rhizodeposition affects microbial regulation of soil N availability. Elevated CO2 stimulated the amounts of root-derived C and N substrates entering the soil, but without specific exudation of amino acids. Enhanced rhizodeposition was accompanied by a proportional increase in root production, suggesting that rhizodeposition under elevated CO2 only increases when root biomass production is stimulated. The increase in rhizodeposition of N under elevated CO2 comprised a significant portion of the plant assimilated N, and was quickly immobilized by microbes upon entering the soil. This shows another pathway by which elevated CO2 may enhance nutrient limitation in low N-input systems. Alternatively, elevated CO2 may alleviate N limitation by stimulating rhizodeposition induced decomposition, leading to the release of N retained in stable SOM pools. This dissertation shows that increased rhizodeposition of C under elevated CO2 may be responsible for sustained plant growth in low nutrient input FACE systems. Since this mechanism did not increase plant tissue N concentrations, and does not contribute to a net gain of ecosystem N, however, it is not expected to offset nutrient limitation under elevated CO2 in the future (i.e. decades to centuries). A third aim of this dissertation was to increase the understanding of plant specific responses to elevated CO2. Therefore, I compared the responses of plants with genetic similarity but contrasting C allocation patterns, so reducing the number of plant traits that can explain a plants’ response to elevated CO2. In addition, C allocation to roots is a key plant trait for explaining differential responses in C and N cycling as it affects both rhizodeposition and nutrient uptake. I showed that agronomic selection has resulted in a morphological tradeoff, where C allocation to organs associated with C assimilation compared to organs associated with nutrient uptake is favoured in modern cultivars. As a result modern cultivars are more likely to increase shoot biomass production under elevated CO2 than their wild relatives in fertilized ecosystems. On the other hand, greater root production and N uptake rates indicate a greater potential for sustained plant growth and soil C sequestration under elevated CO2 for the wild compared to the cultivated genotypes in low N-input systems. These data showed that sink strength is an important trait for controlling plant responses to elevated CO2. In conclusion, elevated CO2 can increase soil C sequestration when sufficient nutrients are available. The extent of the increase however is still unclear in agro-ecosystems, due to a set of management practices that affect soil C decomposition and soil C input. In unfertilized ecosystems, simultaneous increases in N demands of microbes and plants reduce nutrient availability. Increased C allocation to roots under elevated CO2 will benefit nutrient acquisition and C sequestration in low N systems but this mechanism is expected to be transient. Therefore, in natural ecosystems soil C sequestration is likely to be constrained in the future (i.e. decades to centuries) by progressive nutrient limitation.