pastclim: an R package to easily access and use paleoclimatic reconstructions

Abstract

The recent development of continuous paleoclimatic reconstructions covering hundreds of thousands of years paved the way for a large number of studies from disciplines ranging from paleoecology to archaeology, conservation to population genetics, and human evolution to linguistics. Unfortunately, (paleo)climatic data can be challenging to extract and analyze for scholars unfamiliar with such specific file formats. Here we present pastclim, an R package facilitating the access and use of two sets of paleoclimatic reconstructions covering respectively the last 120,000 and 800,000 years. The package contains a set of functions allowing to quickly and easily recover the climate for the whole world or specific areas for time periods of interest, extract data from locations scattered in space and/or time, retrieve time series from individual sites, and easily manage the ice or land coverage. The package can easily be adapted to paleoclimatic reconstructions different from the ones already included, offering a handy platform to include the climate of the past into existing analyses and pipelines.

pastclim 1.2: an R package to easily access and use paleoclimatic reconstructions

“…Leonardi, M. et al 2022. pastclim 1.2: an R package to easily access and use paleoclimatic reconstructions.…”

“…Data are available from the Figshare Repository: https://doi.org/10.6084/m9.figshare.21388545.v3 (Leonardi et al 2022).…”

Following the niche: the differential impact of the last glacial maximum on four European ungulates

“…We would caution against overinterpreting the lack of signal from other mountainous regions: whilst it might derive from their lower altitudes (e.g., the southern Balkans), the inevitably coarse scale of our reconstructions might have prevented us from detecting the effect of mountains with a smaller footprint such as the Pyrenees (which were shown to have acted as a geographic barrier for horses 11,40,41 ). It is of interest that the area covered by the sea of Azov, despite not being excluded from land in the climatic reconstructions 42 is still considered unsuitable for all of our species during the whole period. This suggests that both the reconstructions and the modelling are robust.…”

“…To perform all analyses, we used the R package pastclim v. 1.0 42 to couple each observation from the collapsed datasets to paleoclimatic reconstructions published in 8 by setting dataset = "Beyer2020". These are based on the Hadley CM3 model, include 14 different bioclimatic variables at a spatial resolution of 0.5°, and are available for the whole world every 1000 years until 22 kya and every 2000 years before that date (referred to in the manuscript as "time slices").…”

“…As already mentioned, the four species analyzed show different preferences regarding temperature, habitat, and altitude. Therefore, for the Species Distribution Modelling, we choose five environmental variables that should be able to capture such differences: two measures of temperature (BIO5, maximum temperature of the warmest month, and BIO6, minimum temperature of the coldest month); two variables to help capture habitat differentiation (BIO12, total annual precipitation, and Net Primary Productivity, NPP), and one measure of topography (rugosity 42 ).…”