Data from: Climatic drivers prevail in montane and lowland Odonata latitudinal diversity gradients, but human modification erodes lowland patterns
Data files
Apr 21, 2026 version files 3.03 MB
-
ECOG70066_Rcode.R
22.97 KB
-
Functions.tar
30.72 KB
-
Mountain_dataset.xlsx
1.03 MB
-
Mountain_shape.RDATA
1.94 MB
-
README.md
3.22 KB
Abstract
Latitudinal diversity gradients (LDGs) reflect the interplay of historical, ecological, and evolutionary processes, but these drivers may vary across landforms. Mountains, with steep elevational and climatic gradients, may sustain different diversity dynamics than adjacent lowlands, where human activities are often more intense. Using distributional and phylogenetic data for 732 Odonata species across eastern China, we compared taxonomic and phylogenetic LDGs between mountains and lowlands and identified their environmental determinants. Phylogenetic diversity was quantified using two complementary metrics: mean pairwise distance (MPD), capturing basal-level phylogenetic divergence, and mean nearest taxon distance (MNTD), reflecting tip-level divergence. Odonata species richness and phylogenetic diversity both declined significantly with latitude in mountains and lowlands. The magnitude of these gradients was similar for species richness and tip phylogenetic diversity, but basal phylogenetic diversity declined more steeply in mountains, suggesting stronger constraints on ancient Odonata lineages in montane environments. Random forest analysis revealed that mean annual temperature is the consistent key driver of LDGs in mountains and lowlands, whereas historical climate change had little effect on either. Moreover, temperature seasonality plays a particularly important role in shaping mountain LDGs, while human modification exerts a disproportionate negative effect in lowlands, especially reducing basal phylogenetic diversity. Our results show that Odonata display both taxonomic and phylogenetic LDGs, which are largely similar in mountains and lowlands of eastern China, yet the key factors driving these patterns vary. Montane LDGs primarily reflects temperature seasonality, while lowland LDGs are increasingly shaped by human disturbance superimposed on climatic gradients. Overall, this study highlights the importance of considering landform context when interpreting broad-scale biodiversity patterns, and they emphasize the dual conservation challenge of safeguarding climate-sensitive montane lineages while mitigating the erosion of evolutionary heritage in heavily modified lowland ecosystems.
Dataset DOI: 10.5061/dryad.wpzgmsc2s
Description of the data and file structure
We have submitted our raw data, including two input data files (Mountain_dataset.xlsx and Mountain_shape.RDATA), as well as the R code, comprising two files (ECOG70066_Rcode.R and Functions.tar).
Files and variables
File: ECOG70066_Rcode.R
Description: R_code to loading data, performing data aalysis, and creating figures.
File: Mountain_dataset.xlsx
Description: The raw data file includes two sheets: “Data” (Sheet 1) and “Variables_Abbreviation” (Sheet 2).
The Data sheet provides, for each grid cell, geographic information, environmental variables, biodiversity indices, and a classification indicating whether the grid cell is located in mountainous or lowlands regions based on different buffer criteria (1 km, 5 km, 10 km, 20 km, 30 km, 40 km, 50 km, 60 km).
Detailed descriptions of all environmental variables and biodiversity indices are provided in Variables_Abbreviation sheet.
Variables
- Both unprojected and projected (Birmingham coordinate system) latitude, longitude, and grid cell area are provided.
- A total of 26 environmental variables are included, comprising:
- Current climate variables (19: Bio1-Bio19),
- Historical climate change (1: Velo_sandal),
- Environmental heterogeneity (3: STDEV_Bio1, STDEV_Bio12, and Elevation_range),
- Habitat availability (2: Lotic_Length and Lentic), and
- Human pressure index (1: HMI).
- Biodiversity metrics are provided for each grid cell, including:
- Taxonomic diversity, represented by species richness, and
- Phylogenetic diversity, represented by MPD and MNTD (including mean, standard deviation, standardized effect size (z), rank and p-values).
File: Mountain_shape.RDATA
Description: contains data representing major mountain ranges of China, the mountain range name, and the 50 × 50 km grid cells.
Variables
- Mountain range shape.This dataset contains the polyline layer of major mountain ranges, as provided by Nan et al. (2015). see Reference: Nan, X., Li, A., and Deng, W. 2015. Data set of “Digital Mountain Map of China”. National Tibetan Plateau / Third Pole Environment Data Center.
- Mountain range name. This table provides the names of each mountain range in both Chinese and English.
- Cell grid: We used the Fishnet tool in ArcGIS (version 10.8) to divide the map of China into 50 × 50 km grid cells based on the Birmingham projection. Each grid cell was used as the statistical unit for data analysis.
File: Functions.tar
Description: Some necessary R functions in performing the data analysis. Described in R code.
Code/software
ArcGIS (version 10.8), R software (Version 4.5.1) and Rstudio (2023.12.0 Build 369).
All the work package, see the Rcode.
Access information
Other publicly accessible locations of the data:
- Not applicatable
Data was derived from the following sources:
- Not applicatable
