Animal life on the sea floor has a tremendous influence on the chemical, physical and environmental properties of the surface layer of marine sediment. Most prominent is a process called bioturbation, which comprises the physical mixing of the topmost sediment layer. Without animal activity, the sediment is made up by sequential layers due to perpetual sediment accumulation of new particles settling in the water column. By building burrows, predation or simple locomotion, older sediment material from deeper layers can be excavated and mixed with contemporaneous material from the surface. At the same time fresh material from the surface can enter the deeper and on average, older sediments layer quasi instantly, at least compared to the timescale defined by the slow accumulation of new sediment particles. This has important consequences for palaeoceanographic studies, which collect and measure sediment cores in order to reconstruct local or global climate conditions of the past. Due to bioturbation, climatic events can appear both earlier and later in the sedimentary record and their duration can be severely overestimated. As biologically induced mixing is highly variable and complex, one has to employ a stochastical framework in order to model the effect of mixing on climate signals. In this thesis, different models of bioturbation have been investigated and extended by relaxing constraints in the description of bioturbation and thereby allowing depth and time dependent mixing. The results can be used to estimate potential biases in the sedimentary record and improve our knowledge and understanding of climatic conditions of the Earth system. The order of content is as follows. 1.) LagParSed - A Lagrangian stochastical particle tracking model of diffusive bioturbation. 2.) Organic matter reaction kinetics in bioturbated sediments. 3.) Steady-state age distribution in the bioturbated marine sedimentary record – an Eulerian approach. 4.) Resource-Feedback model of bioturbation and signal preservation