Impacts of nitrogen and phosphorus inputs and elevated atmospheric carbon dioxide on phosphorus dynamics and bioavailability in pasture soils : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Nutrient inputs and climatic conditions can significantly impact soil phosphorus (P) bioavailability and dynamics by influencing chemical, biological, and biochemical processes. This research aimed at investigating the effects of nitrogen (N) and P inputs and elevated atmospheric carbon dioxide (CO2) on soil P bioavailability and dynamics under New Zealand pasture soils. Inorganic P (Pi) and organic P (Po) fractions, together with biological and biochemical properties linked to P cycling, were assessed. In P-deficient soil (4 mg kg-1 Olsen P) with high organic matter content (7.2 %), P addition alone (45 kg P ha-1) or in combination with N (45 kg P ha-1 + 200 kg N ha-1) increased plant growth, total P and N uptake, microbial biomass P, organic anion release, and the depletion of moderately labile Pi and stable Po, whereas N addition alone (200 kg N ha-1) had no impact on plant growth and soil P fractions under legumes (Lupinus angustifolius and Trifolium repens L.) and grasses (Lolium perenne L. and Triticum aestivum L.). In N-limited soil with an Olsen P of 10 mg kg-1, N addition alone (250 kg N ha-1) or in combination with P (250 kg N ha-1 + 50 kg ha-1) increased shoot biomass and shoot P uptake of Italian ryegrass (Lolium multiflorum Lam.) by an average of 1.6-fold compared to the control. In contrast, P addition alone (50 kg P ha-1) had no impact on shoot biomass and shoot N uptake but significantly increased Olsen P and shoot P uptake. Acid and alkaline phosphatase activities increased in summer, especially under N addition treatments, which was attributed to increased plant P demand, alleviation of carbon limitation, and higher temperature. Nitrogen addition alone decreased readily available Pi, labile Pi, labile Po, and moderately labile Pi by 55, 19, 28, and 7 %, respectively, compared to the control treatment. On the other hand, the combined addition of N and P promoted labile Po mineralisation (28 %) and the depletion of labile P fractions but did not mobilise the moderately labile Pi pool compared to P addition alone. In a long-term P fertilisation experiment (0, 188, and 376 kg superphosphate ha-1 yr-1 for more than 65 years) under grazed pastures, soil P availability increased with increasing P inputs, but microbial biomass P was similar under 188 and 376 kg superphosphate ha-1 yr-1. The latter result was attributed to microbial biomass need to maintain nutrient stoichiometry. Alkaline phosphatase activity increased in response to long-term P inputs, whereas acid phosphatase activity decreased. This finding supported the differentiation in origin and nutrient demand between acid and alkaline phosphatase enzymes, where acid phosphatase enzymes are released from plant roots and inhibited by high P availability, whereas alkaline phosphatase enzymes are derived from soil microbes and driven by C availability rather than P. In P-deficient soil (7 mg kg-1 Olsen P), plant biomass, P uptake, and rhizosphere properties did not respond to elevated CO2 (700 ppm) but were significantly impacted by plant species (Lupinus angustifolius, Lolium perenne L., and Triticum aestivum L.). Hence, soil P availability and P fractions distribution were similar under ambient and elevated CO2 due to the negative feedback of P deficiency on photosynthesis. Twenty-two years of CO2 enrichment (500 ppm) under grazed pasture decreased labile and moderately labile Pi by 39 and 15 %, respectively, while promoting the accumulation of moderately labile and stable Po by 26 and 17 %, respectively, compared to ambient CO2. Decrease in soil Pi was attributed to plant P uptake and immobilisation of Pi in the microbial biomass. Accumulation of Po was linked to enhanced biological activity, increased inputs of Po from root detritus, and adsorption onto reactive mineral surfaces of aluminium, iron, and calcium. Plant species investigated in this study and their associated rhizosphere microbes were able to mobilise sparingly available and recalcitrant soil P fractions. This result challenges the recalcitrance of some soil P fractions, which is defined by resistance to extraction and clearly not to action by plants or microbes. Legumes, especially blue lupin, have shown higher ability in releasing organic anions and phosphatase enzymes, which concurred with higher mineralisation of organic P. Under real soil conditions, organic anions played a minor role in Pi acquisition under both legumes and grasses. Nutrient addition experiments and long-term studies revealed that besides P, C and N availability played a critical role in P dynamics and cycling by impacting microbial P transformations and phosphatase enzyme activity. Applications of both N and P in excess of plant requirements cannot accelerate soil P cycling but, in the contrary, can cause soil P accumulation and eutrophication of waterways. Phosphorus deficiency can limit plant ability to grow and acquire P under elevated CO2 conditions. Elevated CO2 can decrease soil P availability under permanent agroecosystems characterized by organic matter accumulation and high availability of free cations (Al, Fe and Ca). Further research is needed to evaluate changes in soil P fractions in response to N and P inputs under N and P co-limited soils. Investigation of the combined effects of elevated CO2, increased temperature, and water stress on soil P availability and dynamics is warranted.