Data from: Climate change could fuel urinary schistosomiasis transmission in Africa and Europe
Data files
Jul 09, 2024 version files 1.18 GB
-
intermediate_mechanistic.zip
507.79 KB
-
occurrence_data.zip
97.56 KB
-
outputs_correlative.zip
241.16 MB
-
outputs_mechanistic.zip
942.95 MB
-
primary_mechanistic.zip
46.77 KB
-
README.md
6.29 KB
Jul 15, 2024 version files 1.18 GB
-
intermediate_mechanistic.zip
507.79 KB
-
occurrence_data.zip
186.25 KB
-
outputs_correlative.zip
241.16 MB
-
outputs_mechanistic.zip
942.95 MB
-
primary_mechanistic.zip
46.77 KB
-
README.md
6.66 KB
Abstract
The freshwater snail Bulinus truncatus is an important intermediate host for urogenital schistosomiasis, a tropical disease affecting over 150 million people. Despite its medical importance, uncertainty remains about its global distribution and the potential impacts of climate change on its future spread. Here, we investigate the distribution of B. truncatus, combining the outputs of correlative and mechanistic modelling methods to fully capitalise on both experimental and occurrence data of the species and to create a more reliable distribution forecast than ever constructed. We constructed ensemble correlative species distribution models using 273 occurrence points collected from different sources and a combination of climatic and (bio)physical environmental variables. Additionally, a mechanistic thermal suitability model was constructed, parameterised by recent life-history data obtained through extensive lab-based snail-temperature experiments and supplemented with an extensive literature review. Our findings reveal that the current suitable habitat for B. truncatus encompasses the Sahel region, the Middle East, and the Mediterranean segment of Africa, stretching from Southern Europe to Mozambique. Regions identified as suitable by both methods generally coincide with areas exhibiting high urogenital schistosomiasis prevalence. Model projections into the future suggest an overall net increase in suitable area of up to 17%. New suitable habitat is in Southern Europe, the Middle East, and large parts of Central Africa, while suitable habitat will be lost in the Sahel region. The change in snail habitat suitability may substantially increase the risk of urogenital schistosomiasis transmission in parts of Africa and Southern Europe while reducing it in the Sahel region.
https://doi.org/10.5061/dryad.98sf7m0s9
This repo contains primary, intermediate, and output data for correlative and mechanistic models for Bulinus truncatus, a schistosomiasis intermedaite host snail. The mechanistic model used is a temperature-dependent intrinsic growth model, calibrated using life history data taken from the literature. The correlative model is an ensemble species distribution model, calibrated using occurrence data from natural history museums and GBIF and bioclimatic variables.
See also the associated publication: van der Deure, T., Maes, T., Huyse, T. & Stensgaard, A-S (2024). Climate change could fuel urinary schistosomiasis transmission in Africa and Europe. Global Change Biology.
Description of the data and file structure
The data is structured in 5 .zip files, which each contain a folder that can be used to reproduce a specific part of our analysis.
primary_mechanistic.zipcontains primary data for the mechanistic model, which are the life history data from several experiments in the literature. In all files, temperature is given in degrees C, proportions as a number between 0 and 1, and other traits have units indicated in the files. Individual-level snail data is from https://www.biorxiv.org/content/10.1101/2024.01.02.573866.abstract. It contains the following files:clutch_size.csvcontains observations of the number of eggs per clutch, linked to individual snailsnumber_of_clutches.csvcontains observations of the number of clutches per snail, linked to individual snailssurvival_data.csvcontains observations of the survival length for individual snails, and from groups of snails from various papers from the literature (for references: see the source column and the supplemantary materials of the publication).traits_data.csvcontains observations of hatching time, hatching success, and maturation time from the literature. For references, see the sourcepaper column and the supplemantary materials of the publication.data_points_plot.csvcontains summary data used in Figure 2 of the associated paper, but not in the analysis.
intermediate_mechanistic.zipcontains samples drawn from the posterior distribution of the life history traits, as well as summaries of the posteriors. These are derived from the life history data using hierarchicical Bayesian models, and are used to calculate the temperature-dependent intrinsic growth rate of Bulinus truncatus. We used a growing degree model to estimate hatching time and maturation time (with two variables T0 and GDD), a quadratic model to estimate egg-laying rate and lifespan (with three variables q, T0, and Tm), and a constant to estimate hatching success (with a single variable q). It has the following files:draws_from_posterior.csvcontains 5000 draws from each posterior distribution, labelled [trait]_[variable], e.g.hatchingtime_T0.- 5 further .csv files (corresponding to each trait) contain the mean, standard deviation, and 2.5% and 97.5% quantile estimate for each variable associated with that trait.
param_estimates.csvcontains the same information as the 5 files named after traits collated into a single file.
occurrence_data.zipcontains three .csv files with all presences, the thinned presences, and background points, namedall_presences.csv,thinned_presences.csv, andbackground_points.csv, respectively. These are used to calibrate the correlative model, and to assess the performance of the mechanistic model. The background points and thinned presences only contain coordinates, while the file will all presences also contains additional information such as the year of publication and identification method.references.docxhas full references for all data points from the published literature.outputs_mechanistic.zipcontains all outputs of the mechanistic model. It contains 5 sub-folders, one for the current climate (namedcurrent) and the others for each combination of socioeconomic pathways (SSP126 and SSP370) and time (middle of the century and end of the century) (named[ssp]_[time]where ssp is either ssp126 or ssp370 and time either 2041-2060 or 2081-2100. Each folder contains the following files:mean.tif, which is a raster file with the mean intrinsic growth rate estimate.median.tif,which is a raster file with the median intrinsic growth rate estimate.prob_present.tif, which is the posterior probability that the intrinsic growth rate is above 0 for any cell.std.tif, which is a raster file with the standard deviation between posterior samples.
The four sub-folders for future climates have four corresponding netCDF (.nc) files, which have the same data, but separately for each global circulation model (GCM), where the Band dimension of the files contains each GCM.
outputs_correlative.zipcontains all outputs from the correlative model. It has the following files and sub-folders:current_set1.tifandcurrent_set2.tif, the mean and median estimates of habitat suitability under current climatic conditions for 2 sets of predictor variables.- four files with the pattern
average_[YY][SSP].tif, with average suitability predictions for each combination of SSP scenario and time. - four files with the pattern
stdev_[YY][SSP].tifwith the standard deviation between GCM scenarios. - A subfolder
by_gcm, which contains result seperately for each GCM, organised in folders named in the same[YY][SSP]pattern, which each contain a file with the suitability prediction for each GCM. - A sub-folder
mopwith results from the mobility-oriented parity (mop) analysis. It contains 5 seperate .tif files for the mop results under current climate (mop_basic_surface_current.tif) and for each combination of time and ssp namedmop_basic_surface_[YY][SSP].tif.
In all the above filenames, [YY] is either 50 (for predictions for 2041-2060) or 90 (for predictions for 2081-2100), and [SSP] is either 126 for SSP126 or 370 for SSP370.
Code
The code to reproduce the analysis is available at https://github.com/tiemvanderdeure/bulinus_truncatus_models.
Version changes
12-07-2024
Added a references file to occurrences.zip, along with minor formatting changes to all_presences.csv such that the sources of occurrence records are easier to identify.
- Maes, Tim et al. (2024), First evidence of a causal link between genetic variation and thermal adaptation in a schistosome host snail, [], Posted-content, https://doi.org/10.1101/2024.01.02.573866
