Too close and too far: quantifying black rhino displacement and location error during research