<i>crestr</i>: an R package to perform probabilistic climate reconstructions from palaeoecological datasets

Abstract

Abstract. Statistical climate reconstruction techniques are fundamental tools to study past climate variability from fossil proxy data. In particular, the methods based on probability density functions (or PDFs) can be used in various environments and with different climate proxies because they rely on elementary calibration data (i.e. modern geolocalised presence data). However, the difficulty of accessing and curating these calibration data and the complexity of interpreting probabilistic results have often limited their use in palaeoclimatological studies. Here, I introduce a new R package (crestr) to apply the PDF-based method CREST (Climate REconstruction SofTware) on diverse palaeoecological datasets and address these problems. crestr includes a globally curated calibration dataset for six common climate proxies (i.e. plants, beetles, chironomids, rodents, foraminifera, and dinoflagellate cysts) associated with an extensive range of climate variables (20 terrestrial and 19 marine variables) that enables its use in most terrestrial and marine environments. Private data collections can also be used instead of, or in combination with, the provided calibration dataset. The package includes a suite of graphical diagnostic tools to represent the data at each step of the reconstruction process and provide insights into the effect of the different modelling assumptions and external factors that underlie a reconstruction. With this R package, the CREST method can now be used in a scriptable environment and thus be more easily integrated with existing workflows. It is hoped that crestr will be used to produce the much-needed quantified climate reconstructions from the many regions where they are currently lacking, despite the availability of suitable fossil records. To support this development, the use of the package is illustrated with a step-by-step replication of a 790 000-year-long mean annual temperature reconstruction based on a pollen record from southeastern Africa.

LegacyClimate 1.0: a dataset of pollen-based climate reconstructions from 2594 Northern Hemisphere sites covering the last 30 kyr and beyond

“…As discussed, all the methods may struggle with some intrinsic characteristics of pollen data and of pollen compilations, including complex species responses, sensitivity to spatial autocorrelation, and limited analogs that may produce poor results in so-called "quantification deserts" (Chevalier, 2019), where fossil pollen is hardly preserved or nearby modern surface pollen samples are missing (Chevalier et al, 2020). However, we designed our datasets so that more methods can be included in our reconstruction scripts (https://doi.org/10.5281/zenodo.7887565; Herzschuh et al, 2023b), such as CREST, an approach that combines presence-only occurrence data from species distribu- tion databases instead of modern pollen samples to estimate the responses of pollen taxa to the climate variable to reconstruct to a climate variable (Chevalier et al, 2014;Chevalier, 2022). CREST is, therefore, more independent from the availability of modern pollen samples.…”

Miocene Climatic Optimum fungal record and plant-based CREST climatic reconstruction from southern McMurdo Sound, Antarctica

“…To reconstruct the regional paleoclimate of Antarctica and better capture uncertainties in climate reconstructions dur-ing the MCO, pollen data from the Antarctic geologic drilling (ANDRILL) program's AND-2A core in the Ross Sea (Warny et al, 2009), and CREST (Climate Reconstruction SofTware), a Bayesian (probability-based) paleoclimate reconstruction technique (Chevalier, 2022a) was used. Five bioclimatic variables were assessed in detail: mean annual temperature (MAT), mean temperature of warmest quarter (MTWQ), mean temperature of coldest quarter (MTCQ), mean annual precipitation (MAP), and precipitation seasonality (CoV × 100).…”

“…Taxa lists and nearest living relatives were extracted from the literature (Warny et al, 2009;Feakins et al, 2012) and used as the input for CREST. CREST reconstructions were carried out in R using the "crestr" package (RStudio Team, 2022;Chevalier, 2022a). Original documentation of the "crestr" R code is available with the package code (https://github.com/mchevalier2/crestr, last access: 28 October 2022).…”

“…CREST requires a minimum of five taxa to work effectively (Chevalier et al, 2014;Chevalier, 2022a). The reconstruction presented herein is based on 13 taxa with a widespread distribution.…”

Ten simple rules to bridge ecology and palaeoecology by publishing outside palaeo-ecological journals

“…This preconception barrier of diverse audiences to "palaeo" methods and the time constraint in engaging with these audiences often leads to the classification of palaeoecology as a separate field rather than a discipline of ecology (Rull, 2014). Progress in effective collaboration through open science (Koren et al, 2022) and computational palaeoecology (e.g., Anderson et al, 2006;Nieto-Lugilde et al, 2021;Chevalier, 2022), as well as adopting approaches that have so far been applied mainly in for ecological studies (e.g., the use of organismal functional traits; Marcisz et al, 2020;Brown et al, 2023) have enhanced the capability to integrate long palaeo-records of microbiota, plants, animals, and abiotic factors with directly observed modern records (data spanning the last 50 years or less; Dillon et al, 2023) (Fig 1). Still, more effort and input are needed from palaeoresearchers (researchers working with palaeoarchives) to integrate palaeoecology within the broader field of ecology sufficiently to routinely include a palaeo-perspective in scientific discussions about the present and future environmental challenges (Camperio et al, 2023).…”