Maxent results

The results section is devided in three sections all containing the results derived with on of the three methods to calculate niche overlap.

performance of ecological niche models

• AUCValues.csv contains the area-under-the-curve values that were compared against those obtained by random sampling within the buffered species area in the ‘main’ maxent workflow
• traits_contribution_maxent.csv contains the contribution of each GIS layer for each species. This is necessarily a sparse matrix because most layers are thrown out (to avoid collinearity) during the variable selection procedure. This spreadsheet is produced by the variable importance workflow.
• mean_traits_contribution_maxent.csv summarizes the results of the previous spreadsheet to obtain the importance of each variable averaged over all species. Also produced by the variable importance workflow.

overlap in niche trait space based on modelled habitat projections

• clustering_nj_MaxEnt_occurences.pdf shows the mean pairwise patristic distance between domesticated and domesticated Ungulates in red and the distance between domesticated Ungulates and wild Ungulates in the blue bars.
• clusters.Maxent.traits.gower.csv contains per species the cluster it belongs to based on niche trait overlap derived from the MaxEnt habitat projections.
• dendrogram.Maxent.traits.gower.pdf shows the clusters plotted on a dendrogram based on traits derived from MaxEnt habitat projections.
• magnitudes.Maxent.csv the dataframe contains information about the importance of the environmental traits in the differentiation of the clusters that contain domesticated Ungulates. The clusters are based on the traits per species derived from the MaxEnt habitat projections.
• maxent_omi.csv contains the distances between the species measured by Gower’s distance on the previously calculated niche traits.
• maxent_omi_hclust.tree phylogenetic tree based on hierarchical agglomerative clustering of the distance data.
• maxent_omi_nj.tree phylogenetic tree based on neighbor-joining of distance data.

overlap in niche trait space derived from raw occurrence data

• clustering_nj_raw_occurences.pdf shows the mean pairwise patristic distance between domesticated and domesticated Ungulates in red and the distance between domesticated Ungulates and wild Ungulates in the blue bars. The distances are based on the niche traits derived from raw occurrence points.
• clusters.raw.occurences.gower.csv contains per species the cluster it belongs to based on niche trait overlap derived from raw occurrence data.
• dendrogram.raw.occurences.gower.pdf shows the clusters plotted on a dendrogram based on traits derived from raw occurrence data.
• magnitudes.raw.csv contains information about the importance of the environmental traits in the differentiation of the clusters that contain domesticated Ungulates. The clusters are based on the traits per species derived from raw occurrence data.
• raw_omi.csv contains the distances between the species measured by Gower’s distance on the previously calculated niche traits.
• raw_omi_hclust.tree phylogenetic tree based on hierarchical agglomerative clustering of the distance data.
• raw_omi_nj.tree phylogenetic tree based on neighbor-joining of distance data.

overlap in niche space derived from modelled habitat projections

• clustering_nj_schoener.pdf shows the mean pairwise patristic distance between domesticated and domesticated Ungulates in red and the distance between domesticated Ungulates and wild Ungulates in the blue bars. The distances are based on the niche space overlap derived from MaxEnt habitat projections.
• clusters.Maxent.Schoener.csv contains per species the cluster it belongs to based on niche overlap derived from the MaxEnt habitat projections.
• dendrogram.Maxent.Schoener.pdf shows the clusters plotted on a dendrogram based on MaxEnt habitat projections.
• overlap.csv contains the overlap between the species niche space measured with Schoener’s D. To create a measure over distance we subtractedthe overlap, i.e. the inverse of the overlap from 1.
• inverse_overlap_nj.tree phylogenetic tree based on hierarchical agglomerative clustering of the distance data.