Skip to main content
Dryad logo

Mammals on mountainsides revisited: trait-based tests of assembly reveal the importance of abiotic filters


Kohli, Brooks; Stevens, Richard; Rickart, Eric; Rowe, Rebecca (2021), Mammals on mountainsides revisited: trait-based tests of assembly reveal the importance of abiotic filters, Dryad, Dataset,


Aim: Mountains provide uniquely informative systems for examining how biodiversity is distributed and identifying the causes of those patterns. Elevational patterns of species richness are well-documented for many taxa but comparatively few studies have investigated patterns in multiple dimensions of biodiversity along mountainsides, which can reveal the underlying processes at play. Here, we use trait-based diversity patterns to determine the role of abiotic filters and competition in the assembly of communities of small mammals across elevation and evaluate the surrogacy of taxonomic, functional, and phylogenetic dimensions of diversity.

Location: Great Basin ecoregion, western North America

Taxon: Rodents and shrews

Methods: The elevational distributions of 34 species of small mammals were determined from comprehensive field surveys conducted in three arid, temperate mountain ranges. Elevation-diversity relationships and community assembly processes were inferred from phylogenetic (PD) and functional diversity (FD) patterns of mean pairwise and mean nearest neighbor distances while accounting for differences in species richness. FD indices were calculated separately for traits related to either abiotic filtering (β-niche traits) or biotic interactions (α-niche traits) to test explicit predictions of the role of each across elevation.

Results: Trait-based tests of processes indicated that abiotic filtering tied to a strong aridity gradient drives the assembly of both low and high-elevation communities. Support for competition was not consistent with theoretical expectations under the stress-dominance hypothesis, species interactions-abiotic stress hypothesis, or guild assembly rule. Mid-elevation peaks in species richness contrasted with overall FD and PD, which generally increased with elevation. PD and total FD were correlated on two of three mountains.

Main conclusions: The functional diversity of small mammal communities in these arid, temperate mountains is most consistent with abiotic filters whereas support for competition is weak. Decomposing FD into traits related to separate assembly processes and examining ecoregional variation in diversity were critical for uncovering the generality of mechanisms. Divergent patterns among dimensions revealed species richness to be a poor surrogate for PD and FD across elevation and reflect the effect of biogeographic and evolutionary history. This first analysis of elevational multi-dimensional diversity gradients for temperate mammals provides a versatile framework for future comparative studies.


Species occurrence data for 34 species of small mammal were obtained through field work by the authors and numerous collaborators. Occurrences have grouped into 100-m elevational bins and presences interpolated between the highest and lowest record of observation.

Species trait data were compiled data from existing databases and the literature for 23 traits.  These were organized under four niche axes: environmental filtering-related traits (ie. beta traits; n = 8), biotic interaction-related traits (ie. alpha-traits; n = 12), life history (n = 2), and body size. 2% of trait values were estimated using regression or genus-level averages.

Phylogenetic relationships and branch lengths (in millions of years) were obtained from the "best dates” supertree of Fritz et al. (2009) pruned to the 34 rodents and shrews in our dataset, with two polytomies resolved using the "multi2di" function in R package 'ape'.

Climate data (mean annual temperature and mean annual total precipitation) were calculated using PRISM data ( extracted from points along each 100-m contour line across the elevational extent of the mountains. An index of aridity was caluclated according to the formula MAP/(MAT+10).

Diversity values were calculated for several indices of phylogenetic and functional diversity using available R packages and standard procedures.

Usage Notes

Files stored here include the complete species trait matrix (Great_Basin_mammal_trait_data_Kohli_etal_JBI.csv) and fully dichotomous phylogeny (GBTree.di_Kohli_etal_JBI.nex) used in all analyses, as well as a summary of species occurrences by elevational bin per mountain (3 MtnName_ElevBins files), the mean climate data (ElevBin_Climate_Kohli_etal_JBI.csv), and calculated diversity metrics and statistics for each elevational bin (DiversityValues_Kohli_etal_JBI.csv).  The Diversity Values file contains columns for Mountain, elevational bin (ElevBin), species richness (Richness), and the calculated index values, SES, and p-value of the SES value for 8 metrics: mean phylogenetic distance (MPD), mean nearest phylogenetic distance (MNPD), and mean functional distance (MFD) and mean nearest neighbor functional distance (MNND) for all traits (.all), beta-traits related to environmental filtering (.ef), and alpha-traits related to biotic interactions (.bi).

For more detailed information, see ReadMe.txt file.


American Society of Mammalogists, Award: ASM Fellowship

National Science Foundation, Award: DEB-1457530