Climate-induced physiological stress drives rainforest mammal population declines
Data files
May 05, 2025 version files 512.89 KB
-
archeri_model_1.R
11.62 KB
-
archeri_model_2.R
11.70 KB
-
archeri_model_3.R
11.70 KB
-
archeri_model_4.R
11.70 KB
-
archeri_model_5.R
11.70 KB
-
archeri_model_6.R
11.70 KB
-
archeri_model_7.R
11.70 KB
-
biophysical_predictors.csv
155.68 KB
-
covs.csv
3.53 KB
-
foliage_data_expanded.csv
15.62 KB
-
grtp_endoR.R
46.11 KB
-
JAGS_code.txt
11.19 KB
-
lemuroides_model_1.R
13.11 KB
-
lemuroides_model_2.R
12.86 KB
-
lemuroides_model_3.R
12.86 KB
-
lemuroides_model_4.R
12.86 KB
-
lemuroides_model_5.R
12.86 KB
-
lemuroides_model_6.R
12.86 KB
-
lemuroides_model_7.R
12.86 KB
-
lrtp_endoR.R
46.05 KB
-
population.csv
56.58 KB
-
README.md
6.02 KB
Abstract
Climate change is a major driver of global biodiversity loss, yet the precise mechanisms linking climate change to population declines remain poorly understood. We developed a novel, broadly applicable framework that integrates biophysical, nutritional and population modelling to capture fundamental physiological constraints on mammalian herbivores and applied it to investigate the causes of declines in ringtail possums of the Australian Wet Tropics (Pseudochirops archeri and Hemibelideus lemuroides). Our approach bridges the gap between mechanistic ("bottom-up") models, which simulate species’ responses based solely on their traits and local microclimates, and the more common ("top-down") statistical models, which infer species’ responses from occurrence or abundance data and standard environmental variables. We quantified population dynamics over a 30-year period by generating species-specific estimates of temperature and water stress, foraging limitations, and linking these with annual monitoring and nutritional quality within an open population model. Our findings demonstrate that climate change has impacted populations through physiological stress, but in a species-specific manner. Both species have experienced population collapses at lower elevations and in low-nutritional sites. For P. archeri, we found evidence that population changes were driven by reduced survival due to overheating and dehydration, alongside diminished recruitment from limited foraging. In contrast, our model suggests that H. lemuroides populations were primarily affected by foraging constraints, emphasising the importance of considering climate-driven limitations on foraging activity in addition to direct physiological stress. These mechanistic insights offer a foundation for targeted conservation strategies to mitigate the impacts of climate pressures on wild populations.
https://doi.org/10.5061/dryad.fxpnvx13n
Description of the data and file structure
File "covs.csv":
- "site": study site ID
- "geology": geological parent materials
- "elev": elevation in m
- "lat": latitude
- "long": longitude
- "K": potassium (cmon_c/kg)
- "N_kje": Kjeldahl nitrogen (%)
- "NA": sodium (cmon_c/kg)
- "P_kje": Kjeldahl phosphorus (%)
- "map": mean annual precipitation (mm)
- "mat": mean annual temperature (ºC)
- "cn": carbon to nitrogen ratio
- "basalt": variable indicating whether a site is on basaltic soils
Missing values in soil parameters are related to unsampled sites. This information was input through the Bayesian hierarchical biogeochemical model developed for the study (see JAGS code file).
File "biophysical_predictors.csv"
- "bio_year": biological year starting in April
- "site": site ID
- "species": ringtail possum species
- "short_ewl": short-term evaporative water loss (g of H2O)
- "long_ewl": long-term evaporative water loss (%)
- "short_forage": short-term foraging performance (%)
- "long_forage": long-term foraging performance (%)
- "short_thermal": short-term thermal stress (ºC hrs)
- "long_thermal": long-term thermal stress (ºC hrs)
Stress/Performance category | Scale | Definition | Units | Variable |
---|---|---|---|---|
Thermal | Long-term | Cumulative hourly stress in a year. Thermal stress was defined as hours when predicted body temperature exceeded a threshold (0.5ºC above the species’ target core temperature), with stress values indicating the amount of heat that surpassed this core temperature threshold. Stress values were summed annually. | ºC·hrs | Thermal dosage |
Thermal | Short-term | Cumulative hourly stress during the highest intensity extreme heatwave recorded in a year. An extreme thermal event was defined as 5 or more consecutive hours of thermal stress, as proposed by A. K. Krockenberger et al. (2012) | ºC·hrs | Thermal intensity |
Water loss | Long-term | Percentage of the year in which the animal is significantly water stressed, with water stress defined as hourly evaporative water loss greater than the 90th percentile of all values (across all sites and years). | % | Long-term evaporative water loss (EWL) |
Water loss | Short-term | Maximum cumulative 4-days water loss | g of H2O | Short-term EWL |
Foraging | Long-term | Percentage of nights in summer with a cumulative foraging time above 80% of the maximum foraging potential. | % | Long-term foraging performance |
Foraging | Short-term | Minimum cumulative 2-days foraging time. Values were scaled using the species’ potential maximum foraging time to offer a comparable metric between species. | % of potential maximum | Short-term foraging performance |
File "foliage_data_expanded"
- "species": tree species name
- "site": site ID
- "rep": replicate number
- "age": leaf age
- "tree_id": tree ID
- "availN": available nitrogen (%)
- "tannins": tannins binding effect
- "DMD": dry matter digestibility (%)
- "total_N": total nitrogen (%)
File "population.csv"
- "site": site ID
- "species": species code (hl = Hemibelideus lemuroides, pa = Pseudochirops archeri)
- "replicate": replicate survey number
- "year": biological year starting in April
- "n_observer": number of observers per survey
- "date": date of survey
- "count": number of different individuals detected
For some surveys, the date and or the number of observers are missing; their values have been replaced with NAs.
Code description
File "archeri_*model_X.R" and "lemuroides_*model_X.R" are the code running the biophysical model for both species. The different numbers correspond to the increasing metabolic cost, with each code modifying the metabolic multiplier in line 86.
File "JAGS code.doc" provides the integrated model that produces the results of the paper taking as input the data provided above.
Files "grtp_*endoR.R" and "lrtp_*endoR.R" are the endotherm function needed to run the species model provided above.