The distortion is occurred due to the difference in ionic radii of dopant and atom in host lattice. In your case, Zn and Zr replacing Fe. The ionic radii of both the dopants are higher than Fe+3 ion, so, with doping the lattice goes expanding. This results in shifting of XRD peak to lower 2 theta.
As noted by Gerhard and Suraj, the XRD peak shifts in the opposite direction of the magnitude difference. A larger dopant shifts the peak to a lower angle and vice versa. This follows Vergard's law well. In solid solution, the lattice parameters shift linearly as a function of the doping level of the host element. Here is another way to interpret your situation. If the Bragg peak does not or does not shift reasonably in the XRD pattern regardless of the doping level, but with only slight intensity changes and nonlinear behavior with the doping level, this is the CN of the cations surrounded by oxygen ions, instead of the average unit cell size change. (number of coordination) can be changed. Alternatively, an asymmetric bond length between atoms can be formed along a specific direction to compensate for energy without changing the lattice constant by changing the crystal field of the system. Possible causes of this include differences in ion size and charge valence, and the creation of vacancy. In this case, you can see big changes whatever the resulting properties are. However, because of the local distortions, it is not easy to interpret what happened in the overall structure using typical XRD measurements. It may be necessary to calculate all distances of atomic bonds through Rietveld fine-tuning, or measure the pair distribution function (PDF) to estimate bond length change and CN change. Please refer to my literature or contact me separately. We will introduce laboratory-based XRD and PDF analysis solutions.