Lebanon biogeography outlined by tree and shrub species distribution pattern

Stephan, Jean 1 ; Hammoud, Youmna1 ; Korban, Melissa1 ; Ferro, Ignacio2

Published Mar 07, 2025 on Dryad. https://doi.org/10.5061/dryad.hmgqnk9vf

Data files

Mar 07, 2025 version files 3.77 MB

cleaned_supportive_information.xlsx
3.76 MB
README.md
7.83 KB

Abstract

This dataset contains environmental study data on various tree and shrub species in Lebanon. The research examines multiple environmental factors influencing species distribution and presence. Occurrence points were collected from field surveys under various studies, covering Lebanon. We used all the presence points and consistent environmental variables called predictors to conduct ecological niche modelling and further use the outcomes of the predicted distribution of tree and shrub species to outline Lebanon’s biogeography. The data referred to as predictors for ecological niche modelling are the environmental factors used for the study, retrieved from open sources. These include climatic, topographic, and biogeographic factors and predictors retrieved from remote sensing data such as cloud coverage, normalized difference vegetation index, and land surface temperature. This information is used to delineate suitability maps for each species and understand the factors driving their distribution. In the next step, species were classified based on their biogeographical appurtenance and grouped into regional chorotypes. The produced heat maps in addition to dissimilarity analyses allowed us to delineate the biogeographical zones and their transition areas in Lebanon. Moreover, this data is being analyzed to study vegetation variation across an elevation gradient.

Description of the data files

The dataset consists of structured tabular data across three sheets originating from three articles:

Stephan et al 2020
Hammoud and Stephan 2022
Stephan and Korban 2025

From the list of the predictors used in each of the three publications, we used those listed in the table below:

Table 1. Predictors: Data source and calculation tools

Code Predictor Source Tool

DEM Digital elevation model (m) Contour line(a) Topo to raster

Aspect Aspect (°) DEM Aspect

Slope Slope (%) DEM Slope

DFS Distance from the sea (m) DEM Fishnet - Near

CC Mean Cloud Coverage (May -July) (%) Cloud Raster(b) Mean Calculation

Tmin Mean of Minimal Temperature of the coldest month (°C) Mean of Minimal Temperature of the coldest month (c) Mean Calculation

Tmax Mean of Maximum Temperature of the hottest month (°C) Mean of Maximum Temperature of the hottest month (c) Mean Calculation

P Cumulative Annual Precipitation (mm) Precipitation Isohyets(d) -

EQ Emberger Quotient Cumulative annual precipitation; Raster Calculator

	Mean Minimal Temperature of the Coldest Month;	

	Mean Maximal Temperature of the Hottest Month

AWC Available water content (mm) Soil maps -

Flow

Flow Accumulation 	DEM	Fill; Flow direction; Flow Accumulation

NDVI Normalized Difference Vegetation Index Mosaic of Landsat 8-Operational Land Imager images Raster Calculator

LST Land Surface Temperature (°C) average for summer months Mosaic of Landsat 8 Images Raster Calculator

LD Land Degradation Land Productivity(e) SDG Land degradation plugin

	Land Cover(e)	

	Soil Organic Carbon(e)

Soil type Geze Soil Geze classification(f) Feature to Raster

(a) The Lebanese Army Geographic Department; (b) National Scientific Research Center (2002 – 2013); (c) WorldClim (1970 – 2000); (d) National Scientific Research Center; (e) Trends.Earth Data, Sustainable Development Goals Plugin.; (f) National Scientific Research Center, Soil Maps.

Each file contains a list of columns including:

Species Latin name (and code)
Geospatial coordinates marking exact species locations in meters (X, Y using Lambert projection, with Deir el Zor as a reference), and decimal degree (global reference). We used the latter for all presence points of the three publications.
Environmental variables (predictors) are listed in Table 1, with the detailed sources of these variables, means of their calculation, and units whenever valid.

The predictors used in our study are among the variables listed in the columns (in bold) of the three papers and are detailed as follows:

FID presence point code

Species_Name species name code

Species species Latin name

X Longitude in meters using Lambert projection

Y Latitude in meters using Lambert projection

Longitude Longitude in decimal degrees, using WG84 projection

Latitude Latitude in decimal degrees, using WG84 projection

Aspect slope direction to the north, in degrees

AWC estimated in mm from soil

depth and texture as extracted from

CNRS soil maps

cumul cumulative average annual precipitations in mm extracted from worldlim (omitted from last model fit)

Curvature inverse length in meters, taken from 10m topographic maps (omitted from last model fit)

DEM digital elevation model taken in meters from 10m topographic maps

Distance (DFS) distance from the sea, estimated in meters from the closest point to the location

Emb_max (Q) Emberger quotient calculated from cumulative annual precipitation (mm) of the highest isohyet, divided by the square of the Mean of Minimum Temperature of

The coldest Month and the square of the Mean Maximum temperature of the warmest month (converted from Kelvin to degrees)

Emb_min Emberger quotient calculated from cumulative annual precipitation (mm) of the lowest isohyet, divided by the square of the Mean of Minimum Temperature of

The coldest Month and the square of the Mean Maximum temperature of the warmest month (converted from Kelvin to degrees); (omitted from the last model fit)

Flow_acc calculated in number of cells, using digital elevation models to determine how much flow is directed to each cell in a raster grid.

Hillshade derived from the digital elevation model through altitude, slope, and aspect, based on the sun-defined position (angle from the north, which is 34 degrees in the case of Lebanon; omitted from the last model fit)

IMI Integrated moisture index derived from hill shade, curvature, flow accumulation, and available water content (omitted from the last model fit)

Max08 (Tmax) Mean maximum temperature of the warmest month in degree Celsius

Mean5_7 (Coverage) Cloud coverage (in percentage of the sky) between May and July

Min02 (Tmin) Mean minimum temperature of the coldest month in degree Celsius

Mother_roc bedrock type derived from the geological map of Lebanon

NDVI Normalized Difference Vegetation Index from mosaic Landsat images

PDIR Potential Direct Incident Radiation (MJ/cm2 year-1) calculated based on latitude, folded aspect (22-degree north is the used azimuth), and slope (extracted from DEM) (omitted from the last model fit)

rain_max_p (P) cumulative average annual precipitations in mm extracted from Atlas climatic du Liban (using the highest isohyet for each presence point)

Slope derived from DEM in percent

Geze_Soil derived from the geological map of Lebanon (Raymond Geze)

LST Land Surface temperature in degrees Celsius

Land Degradation using change in land cover and land use, land productivity, and soil organic carbon data using the Land Degradation Plug-in

Notes

Species data with corresponding observational and measurement records are displayed in rows and sorted by species.

Missing values are replaced with "n/a" to indicate unavailability. These are points where data for the environmental variables are absent.

This dataset is intended for researchers studying environmental influences on species distribution and ecosystem interactions. Users unfamiliar with the dataset should refer to the original study documentation for variable definitions and interpretation.

Sharing/Access Information

Data was derived from the following sources, in which we further detail our methodology:

Stephan, J., Hammoud, Y., Korban, M., & Ferro, I. (2025). Lebanon biogeography outlined by tree and shrub species distribution pattern. Ecology and Evolution (in process).
Stephan, J., Bercachy, C., Bechara, J., Charbel, E., & López-Tirado, J. (2020). Local ecological niche modelling to provide suitability maps for 27 forest tree species in edge conditions. iForest-Biogeosciences and Forestry, 13(3), 230.https://doi.org/10.3832/ifor3331-013
Hammoud, Y., & Stephan, J. (2022). Novelty predictors for shrub (and climbers) ecological niche modeling, based on their successional stage. Ecological Informatics, 71, 101771.https://doi.org/10.1016/j.ecoinf.2022.101771
Stephan, J., & Korban, M. (2025). Ecological niche modelling using MaxEnt for riparian species in a Mediterranean context. Ecological Indicators, 171, 113167. https://doi.org/10.1016/j.ecolind.2025.113167

Code/Software

This dataset is provided in a cleaned Excel format (.xlsx) and can be accessed using any spreadsheet software (e.g., Microsoft Excel, Google Sheets, LibreOffice Calc) or programmatically via Python (pandas) or R (readr).

For further analysis, users may employ Python or R scripts for data processing and visualization. Recommended packages include Pandas, Numpy, and Matplotlib for Python, as well as tidyverse and ggplot2 for R.