Data from: The interplay of traits, phylogeny, and abundance shapes spatial mammal diversity patterns
Data files
Nov 10, 2025 version files 14.77 MB
-
README.md
6.25 KB
-
Supplementary_data1.csv
4.84 KB
-
Supplementary_data2.csv
14.72 MB
-
Supplementary_data3.csv
15.57 KB
-
Supplementary_data4.csv
14.49 KB
-
Supplementary_data5.csv
2.06 KB
Abstract
Terrestrial mammal communities are crucial indicators of the health and sustainability of tropical ecosystems. Consequently, understanding the mechanisms of community assembly and the population dynamics that determine the structure and composition of these communities across spatial and temporal scales is crucial, yet these factors are often overlooked when identifying conservation hotspots. This study aims to evaluate the relative importance of environmental filtering processes in shaping mammal communities at different spatial scales and understand how these processes are influenced by species traits, phylogenetic relationships, and abundances. Ultimately, high mammal diversity areas reflecting underlying assembly mechanisms are identified as conservation hotspots. We analyzed the occurrence data of 44 terrestrial mammal species, collected over six years from 936 camera trap stations in the Indo-Burma biodiversity hotspot, including Upper Myanmar and Xishuangbanna, China, using community hierarchical models. Climate and habitat factors predominantly drive species richness, while traits and phylogenetic relatedness together shape distinct community patterns. Large-bodied mammals with omnivorous guilds prefer stable temperature niches, while reduced precipitation tends to filter communities with wide-ranging mammals. Anthropogenic disturbances and reduced forest cover alter community structures, diminishing the prevalence of carnivorous traits. Phylogenetically clustered niches were observed in the Cervidae and Felidae families, likely driven by shared evolutionary adaptations for temperature tolerance in ungulates and disturbance tolerance in carnivores. Mammal diversity hotspots reflecting distribution and abundance patterns highlight the northwest and southeast ranges of Myanmar as key areas for fostering ecological connectivity with the Far Eastern Himalaya and Dawna-Tanintharyi landscapes. Ultimately, this study presents a novel framework for improving climate- and habitat-based biodiversity models by integrating trait- and phylogeny-driven assembly patterns, while also accounting for population abundances in suitable habitats to define ecologically representative and effective conservation areas.
This dataset accompanies the publication:
Than, K. Z., Hughes, A. C., Wang, L., Zaw, Z., & Quan, R.-C. (2025). The interplay of Traits, Phylogeny, and Abundance shapes Spatial Mammal Diversity Patterns. Global Ecology and Biogeography.
1. Study Overview
The dataset contains camera-trap detection data, environmental and trait covariates, and R scripts used to model mammal richness, occupancy, abundance, and community assembly across Myanmar and southern Yunnan, China.
Objectives:
1. Evaluate the relative importance of ecological assembly processes that sustain terrestrial mammal communities across tropical forest landscapes.
2. Examine trait- and phylogeny-based community responses to environmental and climatic gradients.
3. Generate baseline diversity maps that reflect assembly processes, species’ ecological niches, and population sizes to inform regional conservation planning.
2. Data Summary and File Structure
All datasets are provided in comma-separated (.csv) format, and all analytical codes are in R script format (.R).
File 1 – Supplementary_data1.csv : Sampling Parameters and Weighting
Description: Metadata for each 5 × 5 km sampling grid, including the number of camera traps, sampling occasions, and the weighting factors used to standardize sampling efforts (numbers of camera traps and numbers of sampling occasions) in the Multi-species N-mixture model.
Main columns:
Grid_ID – Unique grid identifier (character)
No_Camera – Number of camera traps deployed (integer)
No_Season – Number of independent sampling occasions (integer)
SWeight – Camera density correction factor (numeric)
CWeight – Camera-day correction factor (numeric)
File 2 – Supplementary_data2.csv : Detection Metadata
Description: Detection records for all species captured by camera traps across six sampling sites. Each row represents an independent detection event.
Main columns:
Grid_ID – 5 × 5 km grid identifier (character)
Camera_ID – Camera trap identifier (character)
Season – Sampling occasion identifier (character)
Latin_Name – Scientific name (based on IUCN taxonomy) (character)
Individual – Unique detection event code (integer)
Site_ID – Sampling site name identifier (character)
File 3 – Supplementary_data3.csv : Species Presence–Absence Matrix
Description: Binary (0/1) presence–absence matrix for 44 mammal species across all sampling grids, derived from the detection metadata.
Notes:
1 = detected in at least one occasion; 0 = not detected.
Used as input for HMSC modeling.
File 4 – Supplementary_data4.csv : Environmental and Habitat Covariates
Description: Scaled covariate dataset used for HMSC modeling, representing habitat, disturbance, and climate factors.
Main variables:
TC – Percent forest cover within grid (Hansen et al., 2013)
FFI – Forest fragmentation index (Ma et al., 2023)
Canopyht – Forest canopy height (Lang et al., 2023)
DEM – Elevation (Caglar et al., 2018)
CBdst – Euclidean distance to the nearest disturbance sources (Esri, 2020; https://livingatlas.arcgis.com/landcover/)
Bio3 – Isothermality in % (WorldClim; Fick and Hijmans, 2017)
Bio19 – Precipitation of the coldest quarter in mm (WorldClim; Fick and Hijmans, 2017)
File 5 – Supplementary_data5.csv : Trait and Phylogenetic Data
Description: Scaled trait dataset for 44 mammal species derived primarily from the PanTHERIA database, with missing values supplemented from the Animal Diversity Web (Myers et al., 2024).
Includes life-history traits used in trait–environment and phylogenetic correlation analyses.
Main traits:
Latin_Name – Scientific name (based on IUCN taxonomy) (character)
BM – Adult body mass (log-transformed) (continuous)
TL – Primary diet category (herbivore, carnivore, omnivore) (categorical)
HR – Mean home-range size (km²) (continuous)
3. Description of Code Files
Supplementary Code 1 – Phylogeny Extraction: Extracts species-level subtree from the global mammal phylogeny. (R: ape, phyloOrchard)
Supplementary Code 2 – Trait Data Extraction: Extracts and scales species-level trait data. (R: dplyr, tidyr)
Supplementary Code 3 – HMSC Model: Runs the spatial hierarchical community model (HMSC). (R: HMSC, R2jags, ape)
Supplementary Code 4 – Model Evaluation: Assesses model explanatory and predictive performance. (R: HMSC, caret)
Supplementary Code 5 – Spatial Predictions: Generates gridded occupancy estimates over the prediction surface based on the best-fitted HMSC model. (R: raster, sf)
Supplementary Code 6 – Multi-species N-mixture Model: Fits MNM and performs bootstrapping for abundance estimation. (R: jagsUI, R2jags, multcompView, caret, clusterGeneration)
Supplementary Code 7 – Diversity Indices: Calculates Shannon and Simpson diversity indices. (R: iNEXT, vegan)
4. Relationships Among Files
Supplementary Data 2 (detection metadata) is processed into Supplementary Data 3 (presence–absence matrix).
Supplementary Data 1 (sampling effort) and Supplementary Data 4 (environmental covariates) are used as predictors in the HMSC and MNM models.
Supplementary Data 5 (traits and phylogeny) is integrated into the HMSC model for trait-based and phylogenetic analyses.
Supplementary Codes 1–7 correspond sequentially to the analytical workflow described in the manuscript.
5. Contact
For questions or data requests, please contact:
Kay Zin Than (PhD)
Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences
Email: kayzin@xtbg.ac.cn, kayzin.pma@gmail.com
Last updated: November 2025
Code/software
R version: 4.2.0 or higher
Required R packages: HMSC, ape, phytools, R2jags, jagsUI, dplyr, tidyr, raster, sf, iNEXT, vegan, caret, multcompView, clusterGeneration
All scripts are annotated and reproducible using standard package functions.
Ensure that working directory paths are set before running.
