Data from: Combining thermal and hydric constraints for spatially predicting the activity suitability of Neotropical Leptodactylid frogs
Data files
Nov 06, 2025 version files 167.08 MB
-
README.md
11.66 KB
-
Senzano_et_al_mechanistic_sdm.zip
167.06 MB
Abstract
The data integrate empirical physiological, behavioral, and environmental parameters to mechanistically assess how temperature and hydration jointly constrain amphibian activity across space and time. We combined experimentally obtained laboratory data — including thermal performance curves for locomotor function, critical thermal limits (CTmin and CTmax), evaporative water loss rates, and dehydration tolerance — with microclimatic datasets. Using these empirical inputs and ready-to-use microclimatic dataset, we parameterized a biophysical modeling workflow that links organismal heat and mass transfer processes with ground-level microclimate conditions. The R scripts provide mechanistic simulations used to estimate hourly body temperatures, evaporative water loss, and cumulative constraints on surface activity time for three Neotropical frog species (Leptodactylus fuscus, L. mystacinus, and L. macrosternum). Model outputs include spatially explicit maps of thermal, hydric, and combined suitability throughout each species’ current distribution and across seasonal phases. This dataset and accompanying R code support these results and provide a detailed framework for integrating physiological and environmental data to model activity suitability under interacting thermal and hydric constraints, offering a reproducible basis for future mechanistic distribution modeling.
Dataset DOI: 10.5061/dryad.02v6wwqhg
Description of the data and file structure
Overview
Amphibians balance their thermal and hydric budgets in response to environmental conditions, with both temperature and hydration jointly constraining activity. Yet, most mechanistic approaches generally emphasize thermal over water limitations. We used a mechanistic modeling framework to assess how interacting thermal and hydric constraints shape the potential activity of three ground-dwelling Neotropical frogs (Leptodactylus fuscus, L. mystacinus, and L. macrosternum) across their geographic ranges. Laboratory-derived physiological data—including thermal performance curves, critical thermal limits (CTmin and CTmax), evaporative water loss rates, skin resistance, and dehydration tolerance—were integrated with microclimatic simulations to parameterize a spatially explicit biophysical model. The model computes organismal heat and water exchanges to estimate hourly body temperatures, evaporative water loss, and nighttime activity time under varying environmental conditions. Model outputs were used to generate thermal, hydric, and joint activity suitability maps across species distributions and seasonal phases. Results revealed that low rather than high temperatures were the main thermal constraint on nocturnal activity, whereas hydric limitations were body size-dependent, being most restrictive for smaller species. This integrative workflow highlights how mechanistic modeling grounded in empirical physiology can reveal species-specific sensitivities to interacting thermal and hydric stressors across space and time.
Files and variables
All data files and variables are annotated within each R script across folders, including variable names and units of measurement. Each step of the processing and analysis pipeline contains comments specifying the origin, structure, and function of variables and dataset used. Where applicable, consistent variable definitions are applied across scripts (provided in-line within each code chunk).
File: Senzano_et_al_mechanistic_sdm.zip
Description:
This repository contains the datasets and R scripts used to generate all mechanistic simulations and analyses reported in the associated publication. The workflow integrates empirical, physiological, and performance data with microclimatic and spatial datasets to predict the joint effects of temperature and hydric constraints on the nighttime activity suitability of three Neotropical Leptodactylus frogs (L. fuscus, L. mystacinus, and L. macrosternum).
Repository structure
├── 01_sdm/ # ensemble species distribution modeling (SDMs)
│ ├── data/ # resources for running ensemble species distribution models (SDMs)
│ │ ├── 00_bio.currSouthAm/ # ready-to-use microclimatic variables (GeoTIFF files)
│ │ ├── extra/ # complementary R functions (contained within R and RDS files)
│ │ └── occ/ # cleaned georeferenced occurrence points for each species (csv files)
│ └── 01_ensemble.R # Runs the ensemble SDM for species distributional ranges (R file)
│
├── 02_mechanistic_act/ # mechanistic modeling
│ ├── 00_microclim/ # ready-to-use microclimatic data from the ‘microclim’ dataset
│ │ ├── 0pctShade/ # Tsub, Tair and RH in 0% shade conditions
│ │ ├── 90pctShade/ # Tsub, Tair and RH in 90% shade conditions (netCDF files)
│ │ ├── SOLR/ # solar radiation (netCDF file)
│ │ └── V1cm/ # wind speed (netCDF file)
│ ├── 01_heat_model_valid/ # biophysical model evaluation (field Tb vs. computed hourly Tb)
│ │ └── data/ # complementary resources for running model evaluation
│ │ ├── agar_models/ # empirical field Tb data collected from agar replicas (text file)
│ │ └── extra/ # auxiliary R functions (R file)
│ ├── 02_database/ # base data for simulations
│ │ ├── ewl_empirical/ # species-specific empirical evaporative water loss (EWL) (text files)
│ │ ├── extra/ # complementary R functions (R file) and South America shape file (RDS file)
│ │ └── micro/ # microclimatic data corrected for canopy classes
│ │ ├── early/ # early season
│ │ │ ├── RH(relat_hum)/ # 1cm above-ground RH (netCDF file)
│ │ │ ├── Soil(subst_temp)/ # substrate (Tsub) and below substrate (Tsub5) temperatures (netCDF files)
│ │ │ ├── SOLR/ # solar radiation (netCDF file)
│ │ │ ├── TA(air_Temp)/ # 1cm Tair (netCDF file)
│ │ │ └── V1cm/ # 1cm wind speed (netCDF file)
│ │ ├── mid/ # mid season
│ │ │ ├── RH(relat_hum)/ # 1cm RH (netCDF file)
│ │ │ ├── Soil(subst_temp)/ # Tsub and Tsub5 temperatures (netCDF files)
│ │ │ ├── SOLR/ # solar radiation (netCDF file)
│ │ │ ├── TA(air_Temp)/ # 1cm Tair (netCDF file)
│ │ │ ├── V1cm/ # 1cm wind speed (netCDF file)
│ │ └── late/ # late season
│ │ ├── RH(relat_hum)/ # 1cm RH (netCDF file)
│ │ ├── Soil(subst_temp)/ # Tsub and Tsub5 temperatures (netCDF files)
│ │ ├── SOLR/ # solar radiation (netCDF file)
│ │ ├── TA(air_Temp)/ # 1cm Tair (netCDF file)
│ │ └── V1cm/ # 1cm wind speed (netCDF file)
│ │
│ ├── 03_Lfuscus/
│ │ └── mechanistic_mod_lf.R # generates mechanistic thermal, hydric, and joint suitability maps # for L. fuscus (R file)
│ ├── 04_Lmystacinus/
│ │ └── mechanistic_mod_lm.R # generates mechanistic thermal, hydric, and joint suitability maps # for L. mystacinus (R file)
│ └── 05_Lmacrosternum/
│ └── mechanistic_mod_lma.R # generates mechanistic thermal, hydric, and joint suitability maps # for L. macrosternum (R file)
└── README.txt
Description of folder structure
The repository is composed of two major components. '01_sdm/' contains the ensemble species distribution modeling (SDMs) workflow, while '02_mechanistic_act/' contains the mechanistic modeling workflow that integrates empirical physiological data and microclimatic inputs to simulate thermal and hydric constraints on frog activity.
The .nc (NetCDF) files contained within the '02_mechanistic_act/00_microclim/' and '02_mechanistic_act/02_database/micro/' directories store spatially explicit hourly microclimatic variables used to parameterize the biophysical models. Each NetCDF file represents hourly climatic data at ~15 km spatial resolution across species-specific distributional ranges.
The variables include air temperature (Tair, 1 cm above ground), substrate temperature (Tsub, surface), subsurface temperature (Tsub5, 5 cm depth), relative humidity (RH, 1 cm above ground), solar radiation (SOLR), and wind speed (V1cm, 1 cm above ground). The microclimatic variables come from the ready-to-use 'microclim' dataset (see Kearney et al., 2014; http://dx.doi.org/10.6084/m9.figshare.878253).
The '00_microclim/' folder contains the existing 'microclim' dataset of Kearney et al. (2014), specifically, for 0% and 90% canopy shading conditions, used to evaluate the biophysical model under field conditions representing sunny and shady scenarios. In turn, the nested directory '02_database/micro/' contains canopy-corrected 'microclim' datasets organized by seasonal phases, 'early/', 'mid/', and 'late/', corresponding respectively to the early (time of season when frogs emerge with the first seasonal rains), mid (frogs' peak activity during the most humid month), and end (when their activity gradually reduces as the rainy season ends) seasonal activity of frogs. Each seasonal folder is further subdivided by climatic variable ('RH/', 'Soil/', 'SOLR/', 'TA/', and 'V1cm/'), each holding NetCDF files with the corresponding hourly microclimatic data.
All simulations were performed in the R language (version 4.1.1) under the Windows 11 operating system. The simulations can be reproduced by running the scripts sequentially, following the folder numbering that indicates their execution order. The workflow integrates various data components, including physiological traits, microclimate data, and species distributional ranges and procedures for model evaluation. Physiological traits comprise mean species-specific values obtained under laboratory conditions. Distributional ranges are derived from α-hull delimitation and species distribution model (SDM) procedures constructed from species' occurrence records. Validation datasets include field-based thermal data from agar replicas and laboratory evaporative water loss (EWL) measurements used to evaluate the predictive accuracy of the biophysical model.
File types and extensions
- NetCDF files (.nc): store gridded hourly microclimatic data.
- GeoTIFF files (.tif): raster outputs from ensemble species distribution models and mechanistic simulations.
- Comma-separated values (.csv): species occurrence records and tabular outputs from SDM models.
- R scripts (.R): code executing analyses, complementary functions, and model simulations.
- R data files (.RDS): serialized R objects (e.g., ready-to-use shapefiles).
- Text files (.txt): field and empirical measurement data.
Code/software
The repository is fully self-contained and structured to run all simulations directly in R without requiring any additional software. All necessary data (raster datasets and physiological information), dependencies, and scripts (including used R-packages and specific functions, along with custom-made functions) are included and organized for complete reproducibility. For detailed information on experimental data collection, physiological trait quantification, the pretreatment of bioclimatic variables and microclimatic datasets, and modeling procedures and assumptions, please refer to the associated publication (Senzano et al., 2025, Ecography), which provides complete methodological and data acquisition details.