Phosphorus (P) is a limiting element in natural and managed ecosystems, with most organic and inorganic P being associated with secondary minerals in acidic soils, and in the form of sparingly soluble calcium phosphates in calcareous ones. For enhancing P availability of these sources, most terrestrial plants establish symbiotic association with arbuscular mycorrhiza fungi (AMF), which in return constitute a major sink of plant photosynthates. While studies on plant P nutrition supplied by soluble inorganic P abound, there is a knowledge gap on the AMF’s ability to mobilise P from primary minerals such as apatite (AP) or orthophosphate (OP) and phytic acid (PA) bound to iron secondary minerals (E.g. goethite -GOE-) in highly weathered soils. Soil P forms are likely differing in their plant acquisition costs, with insoluble organic (E.g. GOE-PA) and inorganic forms (E.g. GOE-OP and AP) are the metabolically most expensive. For AM symbioses it has not yet been shown whether C allocation into the AMF may increase when less accessible soil P forms are available, given that they are available to the AMF at all. I addressed three objectives in the current dissertation. The first objective was to clarify whether an AM plant can take up P from less accessible sources exclusively through the mycorrhizal pathway. The second objective was to explore the carbon (C)–P trading between arbuscular mycorrhizal (AM) plants, along with the ability of this symbiosis to incorporate P derived from differently accessible P sources. Finally, to elucidate the mechanisms involved in the mobilisation of the different P sources by the AM symbiosis was the third objective. For the first and second objectives, I hypothesized (1) AM plants accessing the different P sources will mobilise P in different amounts and rates and (2) the less accessible P sources will require larger photoassimilate investments into the AMF, resulting in differing trading costs per P unit compared to soluble forms. For the third objective I hypothesized (1) P mobilisation will enhance the photosynthetic function in case the AM plant has access to a P source in order to maintain an optimal photosynthetic transfer to the AMF. (2) the AMF development correlates positively with P incorporation rates into the plant tissues, no matter which P source was mobilised; (3) the P mobilisation from the different sources will change the low molecular weight organic acid (LMWOA) profile, with more abundant di/tricarboxylic LMWOA in case of the less accessible P sources. I designed a mesocosms made of a plant compartment containing Solanum lycopersicum plants mycorrhized with the AMF Rhizophagus irregularis DAOM 197198, and a fungal compartment where exclusively the AMF could access and mobilise the different P sources. The C-P trading between an AM plant and its ability to incorporate P derived from differently accessible P sources was tested on fungal compartments containing OP, PA, AP, GOE-PA and GOE-OP, while the identification of mechanisms was checked for OP, PA, GOE-OP and GOE-PA. Over different time course experiments, I measured plant P stocks and the C budget to access the P sources was estimated by measuring the respired CO2, the organic C, the phospholipid fatty acid (PLFA) 16:1ω5c (mycelium biomass) and the neutral lipid fatty acid (NLFA) 16:1ω5c (mycelial energy storage) at the fungal compartment. In order to determine the mechanisms, photosynthetic capacity and mycorrhization rates were measured at the plant compartment, while LMWOA and PLFA microbial biomarkers were monitored at the fungal compartment.All AM plants incorporated P derived from all five sources through the mycorrhizal pathway. They did this at different rates and differing photosynthetic investment costs per unit of P incorporated, yielding differing C:P trading rates. Arbuscular mycorrhizal plants mobilising PA, AP, GOE-OP and GOE-PA caused a larger mycelium infrastructure (PLFA) and a higher energy storage (NLFA) as compared to the OP. All AM plants with access to a P source preferentially stored P in their shoots. This drove up photosynthetic capacity and was accompanied by an improved photosynthetic P use efficiency. Phosphorus incorporation in AM plants correlated with abundance of AMF and bacterial biomarkers in case of PA and GOE-PA, but not in case of GOE-OP. I found di/tricarboxylic LMWOA in treatments containing PA, GOE-PA and GOE-OP before and during the P incorporation, pointing to a ligand exchange mechanism to mobilise the P.The different C investments into P allocation from differently accessible sources suggests a broad nexus between P mining strategies, resource partitioning in soil, and the amounts of C accumulated in terrestrial soils. Furthermore, the current dissertation provides the first evidence on the ability of an AM plant to incorporate P derived from an organic source bound to a secondary mineral (GOE-PA), opening the possibility to access the set of different P forms in soils, and thus reduce the application of fertilizers based on phosphate rock.