A continuously growing world population with a projected size of more than 9 billion inhabitants in the year 2040 requires huge efforts in food production while concurrently avoiding adverse side effects such as the use of pesticides or fertilizers. Among them phosphorous (P) is an important mineral fertilizer for which only few renewable sources exist and which is becoming increasingly scarce. Therefore, methods to reduce P fertilization or enhance fertilization efficiency are urgently needed. One idea is to look how plants in natural ecosystems cope with the problem of nutrient limitation. A strategy, found in almost all plant species is interaction with mycorrhizal fungi. Plants usually deliver carbohydrates (C) to the fungi and get nutrients, like phosphorous (P), in exchange. In natural ecosystems, plants usually interact with multiple fungi which perform differently in their P delivery. However, in agro-ecosystems not all these fungi are helpful. Fungi which are carbon demanding but deliver just few P, might even result in lower plant growth. Therefore a deep knowledge of the mechanisms driving the P and C dynamics is necessary. This can be gained by a computer simulation model which is possible to examine the influence of different nutrient exchange strategies in detail and make prediction how they perform. In this PhD thesis, a spatially explicit simulation model of arbuscular mycorrhizal fungi (AMF) was developed and specific laboratory experiments have been conducted and used for model calibration. This model has been used to evaluate the performance of different nutrient exchange strategies by the emerging maximum achievable fungal biomass, the C uptake rate from the plant and the P delivery rate to the plant. On this basis, three functional types could be identified: parasitic type, intermediate type, mutualistic type. In further steps these functional types have been used to investigate their performance to smooth temporal P pulses (i.e., by transforming them into a continuous P flux delivered to the plant) and to take up spatially heterogeneously distributed P. In both cases, the mutualistic type was found to perform worst and parasitic type best. Two key mechanisms for efficient resource use in spatiotemporally heterogeneous environments could be identified. By the ability of quick fungal biomass growth, AMF can efficiently explore space and store P inside the fungal mycelium. By the creation of spores that do not need C for 6 maintenance, AMF can use the saved C to grow new hypha for further spatial exploration. Through these two mechanisms AMF are able to adapt their mycelium to the spatial and temporal conditions of the P distribution and thus have the potential to largely enhance Puse efficiency. This finally might reduce the application of P fertilizers.