Many biological materials exhibit outstanding mechanical properties that combine stiffness with high toughness and damage tolerance. One example is the skeleton of deep-sea sponges, which shows greatly increased fracture toughness due to soft interlayers (ILs) within a brittle matrix material. These crack arresting properties are explained by the so-called material inhomogeneity effect. Therein, a reduction of crack driving force (CDF) is attributed to the noticeable differences in Young's modulus and yield stress between the constituents. Although comparable structures have already been mimicked using metals or ceramics, a detailed study has not yet been conducted for polymers. Within the scope of this thesis, the material inhomogeneity effect was replicated in bio-inspired composites made of polypropylene-based materials. Different layer architectures were explored, which can generally be divided into microlayer composites, where layers are only a few microns thick, and multilayer composites, where layer sizes are in the sub-millimeter range. The objective is to increase fracture toughness and damage tolerance by using soft ILs, while avoiding stiffness reduction due to the soft material. To enable a thorough investigation of stiffness and toughness in such complex structures, targeted modifications were introduced to existing procedures from elastic plastic fracture mechanics, material testing and simulation. Using these adapted methods, the crucial roles of material selection, defect size, layer thickness and crack re-initiation after arrest were explored in detail. For example, the optimal layer size for microlayer composites was determined as the size of the largest inherent defects. With such an architecture, Charpy impact strength of a brittle matrix material could be increased by more than 11 times. Fracture energy under tensile loading was also improved 33-fold. Unfortunately, the enormous benefits to toughness and flaw tolerance were also accompanied by decreases in stiffness by up to 90%. For applications where stiffness is of high priority, multilayer composites are a promising alternative. Specimen stiffness of these composites can be preserved better due to larger matrix layer sizes. More specifically, the largest improvement to fracture toughness could be measured at 3.86 times the matrix toughness, while 75% of stiffness could be maintained. The best solution in terms of stiffness only sacrificed 6% of the matrix values, while fracture toughness could be increased by a factor of 2.81. Ultimately, guidelines for the design of these optimal structures were deduced, which are intended for later use in engineering applications.