## Locally optimal paths that minimally deviate from the notion of a dominant minimal path and that were derived from the initial condition shown in Fig 5A.

(A) A minimal path (corresponds to the locally optimal landscape shown in Fig 5B) that does not attain the maximal mosaic richness index at the locally optimal landscape (see Fig 8B, solid), hence this minimal path is not a dominant minimal path. B) A locally optimal path (corresponds to the locally optimal landscape shown in Fig 5D) that can be excluded from the list of viable options due to poor relative species richness sustained at each intermediate landscape along the path (see Fig 8B, dotted). (C) A locally optimal path (corresponds to the locally optimal landscape shown in Fig 5F) that supports maximal diversity at each intermediate landscape along the path, relative to all other locally optimal paths that end in a globally optimal landscape (see Fig 8B, dashed), but is not a minimal path. The paths in (A) and (C) are both viable options, but is the end result of a global optima in (C) necessarily better than the local optima that requires fewer parcel exchanges (shown in (A)), which does not reach a global optima?