Figure 2. The difference between the OMI analysis, K-select analysis and WitOMI calculations.

(A) OMI analyses performed on three hypothetical subsets (K₁, K₂, and K₃) and two species (j₁, j₂). The three positions of the two species niches with their corresponding minimum convex polygon (i.e., niche breadth) are not comparable across subsets (K₁, K₂, and K₃) because ordination is performed for each subset, creating new origins, G₁, G₂, and G₃ (i.e., equivalent to the average habitat conditions used by the community). (B) Separate K-select analyses performed for each species, j₁ and j₂, in the three subsets, K₁, K₂, and K₃. The resulting niches for each subset of the two species are not comparable because the origins O₁ and O₂ represent the average habitat used by the species j₁ and j₂, respectively. (C) Species’ niche position and breadth analyzed with the OMI analysis. WitOMI, further decompose the species niche into subniches (j₁ K₁, j₁ K₂, j₁ K₃ and j₂ K₁, j₂ K₂, j₂ K₃ for j₁ and j₂, respectively) and indexes can be calculated from G, WitOMIG. The black dots (G₁, G₂, and G₃), representing the average subset used by one assemblage, are used to calculate subniche indexes, WitOMIG_K.