The climatic space of European pollen taxa

Description

Pollen data are widely used to reconstruct past climate changes, using relationships between modern pollen abundance in surface samples and climate at the surface sample sites. Visualisation of these data in multi-dimensional climate space provides an important way to establish that pollen taxon abundances are well-behaved before using them in climate reconstructions, but visualisation can also be helpful for ecological interpretation of the pollen diagrams. Here we present data created using Generalized Additive Models (GAMs) on the distribution of 195 European pollen taxa in climate space defined by seasonal temperature, as defined by the mean temperature of the coldest month (MTCO) and growing degree days above a baseline of 0°C (GDD0), and an annual moisture index (MI) expressed as the ratio of annual precipitation to annual potential evapotranspiration. These models can be used to explore the realised climate niche of individual pollen taxa and to build statistical models for climate reconstruction.

Resource Type:	Dataset
Creators:	Wei, Dongyang ORCID: https://orcid.org/0000-0003-0384-4340, Harrison, Sandy ORCID: https://orcid.org/0000-0001-5687-1903 and Prentice, Iain Colin ORCID: https://orcid.org/0000-0002-1296-6764
Rights-holders:	University of Reading, Imperial College London
Data Publisher:	University of Reading
Publication Year:	2019
Data last accessed:	3 November 2024
DOI:	https://doi.org/10.17864/1947.204
Metadata Record URL:	https://researchdata.reading.ac.uk/id/eprint/204
Organisational units:	Science > School of Archaeology, Geography and Environmental Science > Department of Geography and Environmental Science
Participating Organisations:	University of Reading, Imperial College London
Keywords:	pollen taxa, climatic space, Generalized Additive Models
Rights:	Creative Commons Attribution 4.0 International
Data Availability:	OPEN
Additional details Additional details Project Name: Global Change 2.0: Unlocking the past for a clearer future Funders: European Commission (694481) Collection period: From January 2018 To April 2019 Geographic coverage: Europe, Middle East, northern Africa, northern Eurasia Data collection method: The modern pollen dataset consists of records from 6458 terrestrial sites from Europe, the Middle East and Eurasia. Climatological values (1961-1990) of mean monthly temperature, precipitation, and fractional sunshine hours were obtained from the CRU CL v2.0 gridded dataset. Data processing and preparation activities: Geographically weighted regression in ArcGIS was used to correct for elevation differences between the CRU grids and the pollen sites, using a fixed bandwidth kernel of 1.06° (~140km). The climate of each pollen site was estimated based on its longitude, latitude, and elevation. MTCO was taken directly from the GWR regression; GDD0 was estimated from daily data using a mean-conserving interpolation of the monthly mean temperatures. MI was calculated for each pollen site using SPLASH v1.0, based on daily values of precipitation, temperature and sunshine hours obtained using a mean-conserving interpolation of the monthly values of each. The GAMs were implemented with the mgcv R package. MI was square root transformed, since differences between MI values at the dry end have a bigger effect than differences at the wet end on vegetation. Logistic models were constructed. Interaction terms were not included, because we assume that each bioclimatic variable has an independent influence on taxa distribution. The fitted response surfaces show pollen taxon abundance in climate space. Convex hulls, implemented using the alphahull package in R, were used to delineate the area with samples and avoid representing parts of the fitted surface that are not closely constrained by data. We sample the resultant GAMs at regular intervals across the range spanned by each climate variable to derive estimates of taxon abundance, sampling each GAM at increments of 0.6°C for MTCO, 100 degree days for GDD0 and 0.03 for √MI (unitless). Depositing User: Dongyang Wei Date Deposited: 03 May 2019 11:59 Last Modified: 29 Jul 2023 04:18