Abstract

The first detailed reconstruction of the terrestrial paleoclimate development of the UK Neogene (?Langhian to Piacenzian) is presented. The paleoclimate data are derived from the paleobotanical record using two probability-based reconstruction techniques CREST (Climate REconstruction SofTware) (Chevalier et al. 2014) and CRACLE (Climate Reconstruction Analysis using Coexistence Likelihood Estimation) (Harbert & Nixon 2015) that use Bayesian and likelihood estimation probability respectively. The results of these reconstructions are presented alongside reconstructions using the widely-applied Co-existence Approach (CA) (Utescher et al. 2014) for comparison. While all three techniques use the climate requirements of their Nearest Living Relatives as the basis of their reconstruction, they use different database observations. CREST and CRACLE use the GBIF (Global Biodiverstiy Information Facility) (GBIF, 2021) as well as WorldClim inputs for the 19 bioclimate variables used by BIOCLIM (http://www.worldclim.org/bioclim). Meanwhile, the CA uses the Palaeoflora database, meaning the input for the three models is different. The reconstructions for the UK Neogene palaeoclimate come from 4 localities (12 samples total) spanning the Middle Miocene (Langhian) to Pliocene (Piacenzian): Trwyn y Parc, Anglesey (Middle Miocene), Brassington Formation, Derbyshire (Serravallian-Tortonian), Coralline Crag Formation (latest Zanclean-earliest Piacenzian) and Red Crag Formation (Piacenzian-Gelasian) of southeast England. We present CREST and CRACLE reconstructions of 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). The CA does not reconstruct MTWQ, MTCQ or precipitation seasonality. Instead, the CA reconstructs Warmest Month Mean Temperature (WMMT) and Coldest Month Mean Temperature (CMMT). The proportion of rainfall falling in the wettest months of the year (RMPwet(%)) was used as a proxy for precipitation seasonality following the methodology of Jacques et al. (2011) and Utescher et al. (2015). The CREST R-code output provides 0.5 and 0.95 (2-σ) uncertainties as well as an optimum and mean for each variable. The CRACLE R-code output provides both parametric and non-parametric joint likelihoods (P-CRACLE and N-CRACLE) with 0.95 (2-σ) uncertainties and a mean that is based on P-CRACLE. The CA generates a minimum and maximum likelihood which together comprise the coexistence interval. The Neogene climate reconstruction of the UK shows a cooling trend from the Langhian to the Pliocene-Pleistocene boundary. CREST and CRACLE produce trends and values consistent with Co-existence Approach data with 0.95 uncertainties overlapping with the CA coexistence interval.

File Descriptions

Table S1 displays the complete reconstruction for the UK Neogene using CREST, CRACLE and the Co-existence Approach.
Table S2 displays detailed site information including: modern and paleo latitude and longitude, dating technique, modern climatology and fossil assemblage diversity (number of fossil taxa versus number of NLRs used for climate reconstruction). Modern climatology has been included to serve as a comparison to the reconstructed Neogene climate. This data has been extracted from WorldClim 2.1 (Fick & Hijmans, 2017).
Data Set S1 contains the list of fossil spore and pollen taxa per site and associated Nearest Living Relatives (NLRs), where identifiable, used as the input for CREST, CRACLE and the Co-existence Approach. Relic taxa are included and highlighted in red.
Data Set S2 is included to show the effect relic taxa have on paleoclimate reconstructions. The relic taxa are removed following the protocol of Utescher et al. (2014) whereby known relic taxa are removed from analyses to avoid biased reconstructions. Relic taxa removed from analyses include Cathaya, Cryptomeria, Pinus sylvestris and Sciadopitys when present.
Data Set S3 is included to show the effects of removing family-level identifications in CRACLE reconstructions. Removing families is shown to generate a less informative reconstruction. Including both genera- and family-level classifications of NLR (Nearest Living Relative) is recommended, however we suggest identifying NLRs (Nearest Living Relatives) to genera-level wherever possible.

Paleoclimate reconstructions appear to be unavoidable steps in the future climate change understanding and especially the local calibration of the paleoclimate proxies. The Mongolian Plateau in particular, is a scarcely studied area. Here we present a latitudinal transect from the southern Siberian Baikal area to the Mongolian part of the Gobi desert: the New Mongolian-Siberian DataBase (NMSDB). The 49 surface samples presented in this dataset are from different types: moss polsters, surface soil samples and lacustrine top-cores. Two paleoclimate proxies have been carried here: pollen analysis and biomarkers (glycerol dialkyl glycerol tetraethers, GDGTs). The actual bioclimate parameters of each sample sites are derived from the ASTER data (NASA, 2014) for the elevation (m a.s.l.) and the WorldClim2.0 interpolated climate database (Fick et Hijmans, 2017) for the climate parameters. Fick, Stephen E., et Robert J. Hijmans. 2017. ' WorldClim 2: New 1-Km Spatial Resolution Climate Surfaces for Global Land Areas: NEW CLIMATE SURFACES FOR GLOBAL LAND AREAS ». International Journal of Climatology 37 (12): 4302‑15. https://doi.org/10/gb2jnq. NASA JPL. 2014. ' ASTER Global Emissivity Dataset, 100-meter, HDF5 ». NASA EOSDIS Land Processes DAAC. https://doi.org/10.5067/COMMUNITY/ASTER_GED/AG100.003.

This data publication is supplementary to a study on the climatic controls on leaf wax hydrogen isotopes, by Gaviria-Lugo et al. (2023). The dataset contains hydrogen isotope ratios from leaf wax n-alkanes (δ2Hwax) taken from soils, river sediments and marine surface sediments along a climatic gradient from hyperarid to humid in Chile. In addition, for each sampling site the hydrogen isotope ratios from precipitation (δ2Hpre) from the grids produced by the Online Isotopes in Precipitation Calculator (OIPC) (Bowen and Revenaugh, 2003). Furthermore, for each sampling site we report mean annual data of precipitation, actual evapotranspiration, relative humidity, and soil moisture, all derived from TerraClimate (Abatzoglou et al., 2018). Also provide data of mean annual temperature and the annual average of maximum daily temperature derived from WorldClim (Fick and Hijmans, 2017). As a final climatic parameter, we also derived data of aridity index from the Consultative Group of the International Agricultural Research Consortium for Spatial Information (CGIARCSI) (Trabucco and Zomer, 2022). In addition to climatic variables, for each site we include land cover fractions of trees, shrubs, grasses, crops, and barren land. These land cover fractions were obtained from Collection 2 of the Copernicus Global Land Cover layers (Buchhorn et al., 2020) via Google Earth Engine. For further comparison here we provide δ2Hwax compiled from 26 publications (see references) that reported both the n-C29 and n-C31 n-alkanes homologues from soils and lake sediments. For each sampling site of the global compilation, we provide δ2Hpre and the same climatic and land cover parameters as for the Chilean data (i.e., precipitation, actual evapotranspiration, relative humidity, soil moisture, aridity index, temperature, fraction of trees, fraction of grasses, etc.), using the same sources. The data is provided here as one single .xlsx file containing 9 data sheets, but also as 9 individual .csv files, to be accessed using the file format of preference. Additionally, 5 supplementary figures that accompany the publication Gaviria-Lugo et al. (2023) are provided in one single .pdf file. The samples taken for this study were assigned International Geo Sample Numbers (IGSNs), which are included in the provided tables S4, S5 and S6.

This dataset consists of combined field transect data and data from a provenance trial. Geographic coordinates and selected Worldclim climatic parameters (mean annual temperature, mean temperature of the warmest quarter and mean annual precipitation) for each sampling site or the original provenance site.are given. For each sampling site we give mean needle length, width, and stomatal traits for each needle surface (number of stomatal rows and stomatal density) as well as the projected stomatal density..