Data from: Prolonged drought on Rapa Nui during the decline of megalithic monument construction
Data files
Dec 16, 2025 version files 110.21 KB
-
Data.xlsx
73.39 KB
-
Rano_Aroi_Age_Model.csv
16.70 KB
-
Rano_Aroi_Chain_Length_Data.csv
3.09 KB
-
Rano_Aroi_Hydrogen_Isotopes.csv
2.07 KB
-
Rano_Aroi_Input_File_for_BACON.csv
476 B
-
Rano_Kao_Chain_Length_Data.csv
2.72 KB
-
Rano_Kao_Hydrogen_Isotopes.csv
3.30 KB
-
README.md
8.47 KB
Abstract
Drought may have contributed to changes in human geography on Rapa Nui in the late 16th and early 17th centuries, but there are currently no continuous paleoclimate records that provide direct, quantitative insight into local hydroclimate conditions during this time period. Here, we present new evidence for a multicentury rainfall deficit on Rapa Nui beginning at approximately 1550 CE, based on two independent reconstructions of hydrogen isotopes in rainfall inferred from hydrogen isotopes in leaf waxes preserved in wetland sediments. Chain-length distribution data suggests that known vegetation changes on Rapa Nui (e.g., palm deforestation) did not significantly influence the sources of leaf waxes for either wetland. Therefore, we assume that leaf wax hydrogen isotopes reflect isotopes in precipitation at a constant offset through time. We use instrumental data from the Global Network of Isotopes in Precipitation (GNIP) to show that there is a significant relationship between isotopes in precipitation over Rapa Nui and mean annual rainfall amount. We then apply the relationship between these two variables to estimate the magnitude of drought at 1550 CE, inferred from our leaf wax records. We estimate that the period 1550 CE to 1720 CE on Rapa Nui saw a 600-800 millimeter decline in average annual rainfall relative to the previous three centuries. This drought is of greater magnitude than the drying observed in recent decades and likely posed significant challenges to the 16th-17th century inhabitants of Rapa Nui.
Description of the data and file structure
Data.xlsx contains all leaf wax hydrogen isotope and chain-length distribution data files for sediment cores from Rano Aroi wetland and Rano Kao crater lake. All sheets are separately provided as CSVs under the same titles. The variable descriptions below are broken up by dataset.
Variable Descriptions
Data file: Rano_Kao_Hydrogen_Isotopes.csv
Variable: Depth (cm)
Description: Depth of each sample in core KAO08-03. Depths are listed as the midpoint between the top and bottom of the sampling interval.
Variable: Year (CE)
Description: Modeled age of the sediment at each sample depth. See the Sharing/Access Information section for more information on the KAO08-03 age-depth model.
Variable: Max Age (CE)
Description: Maximum age of the sediment at each depth, defined as the upper 95 % confidence interval of the age-depth model for KAO08-03.
Variable: Min Age (CE)
Description: Minimum age of the sediment at each depth, defined as the lower 95 % confidence interval of the age-depth model for KAO08-03.
Variable: n-Alkanoic acid chainlength
Description: Number of carbons in the n-alkanoic acids measured for hydrogen isotope composition
Variable: dDwax (‰ VSMOW)
Description: Hydrogen isotope composition of n-alkanoic acids in each sample, averaged across duplicate or triplicate measurements where appropriate. Values have been converted to the VSMOW scale and adjusted to account for the addition of a methyl group of known isotopic composition during the methylation procedure.
Variable: 1s analytical uncertainty dDwax (‰)
Description: Pooled standard deviation from all replicate measurements of a given chainlength, weighted by the number of replicates (n-1) for each sample. Samples run on the same reactor will have the same analytical uncertainty.
Variable: number of replicates
Description: Number of replicates per sample within the range of accepted peak sizes (> 10Vs, < 150Vs)
Data file: Rano_Aroi_Hydrogen_Isotopes.csv
Variable: Depth (cm)
Description: Depth of each sample in core RAP-01-18 Drive 1. Depths are listed as the middle point between the top and bottom of the sampling interval.
Variable: Year (CE)
Description: Modeled age of the sediment at each sample depth. See the "Rano*_*Aroi_Age_Model.csv" section for more information on the RAP-01-18 Drive 1 age-depth model.
Variable: n-Alkanoic acid chainlength
Description: Number of carbons in the n-alkanoic acids measured for hydrogen isotope composition
Variable: dDwax (‰ VSMOW)
Description: Hydrogen isotope composition of n-alkanoic acids in each sample, averaged across duplicate or triplicate measurements where appropriate. Values have been converted to the VSMOW scale and adjusted to account for the addition of a methyl group of known isotopic composition during the methylation procedure.
Variable: 1s analytical uncertainty dDwax (‰)
Description: Pooled standard deviation from all replicate measurements of a given chainlength, weighted by the number of replicates (n-1) for each sample. Samples run on the same reactor will have the same analytical uncertainty.
Variable: number of replicates
Description: Number of replicates per sample within the range of accepted peak sizes (> 10Vs, < 150Vs)
Note: Maximum and minimum age ranges for each depth in Rano Aroi are available in "Rano_Aroi_Age_Model.csv" or in the Data.xlsx spreadsheet.
Data file: Rano_Kao_Chain_Length_Data.csv
Variable: Depth (cm)
Description: Depth of each sample in core KAO08-03. Depths are listed as the middle point between the top and bottom of the sampling interval.
Variable: Year (CE)
Description: Modeled age of the sediment at each sample depth. See the Sharing/Access Information section for more information on the KAO08-03 age-depth model.
Variable: Max Age (CE)
Description: Maximum age of the sediment at each depth, defined as the upper 95 % confidence interval of the age-depth model for KAO08-03.
Variable: Min Age (CE)
Description: Minimum age of the sediment at each depth, defined as the lower 95 % confidence interval of the age-depth model for KAO08-03.
Variable: Average Chain Length (ACL) 26-32
Description: Average chain length for n-alkanoic acids with 26-32 carbons
Variable: CXX fractional abundance
Description: Fractional abundance of n-alkanoic acids with XX number of carbons, relative to the total abundance of waxes C20-C32
Data file: Rano_Aroi_Chain_Length_Data.csv
Variable: Depth (cm)
Description: Depth of each sample in core RAP-01-18 Drive 1. Depths are listed as the middle point between the top and bottom of the sampling interval.
Variable: Year (CE)
Description: Modeled age of the sediment at each sample depth. See the "Rano*_*Aroi_Age_Model.csv" section for more information on the RAP-01-18 Drive 1 age-depth model.
Variable: Average Chain Length (ACL) 26-32
Description: Average chain length for n-alkanoic acids with 26-32 carbons
Variable: CXX fractional abundance
Description: Fractional abundance of n-alkanoic acids with XX number of carbons, relative to the total abundance of waxes C20-C32
Note: Maximum and minimum age ranges for each depth in Rano Aroi are available in "Rano_Aroi_Age_Model.csv" or in the Data.xlsx spreadsheet.
Data file: Rano_Aroi_Input_File_for_BACON.csv
Variable: age
Description: Radiocarbon (14C) age of dated materials (terrestrial macrofossils or pollen extracts) in years before present (yrs BP). "Present" is defined as 1950.
Variable: error
Description: Uncertainty in radiocarbon age (years).
Variable: depth
Description: Depth of each sample in core RAP-01-18 Drive 1. Depths are listed as the middle point between the top and bottom of the sampling interval.
Variable: cc
Description: Radiocarbon calibration curve applied to each sample in BACON. A cc of 3 indicates the use of the Southern Hemisphere calibration SHCal20. A cc of 0 indicates that the sample is from the postbomb period.
Note: In Excel, this sheet contains the prior settings used when running the age-depth modeling software BACON for Rano Aroi core RAP-01-18. This information has been removed in the associated CSV file.
Dataset: Rano_Aroi_Age_Model.csv
Note: The age-depth model for Rano Kao core KAO08-03 was generated using CLAM.R 2.2 with smooth spline interpolation and was previously published in Rull et al., 2018. Please see the Sharing/Access Information section for a full citation. The age model for Rano Aroi core RAP-01-18 Drive 1 was generated using the Bayesian depth modeling software BACON and is presented for the first time in this study.
Variable: depth (cm)
Description: Sediment depth in core RAP-01-18 Drive 1.
Variable: min 95 % (yrs BP)
Description: Minimum modeled age of each depth interval in yrs BP, defined as the lower 95 % confidence interval.
Variable: max 95 % (yrs BP)
Description: Maximum modeled age of each depth interval in yrs BP, defined as the upper 95 % confidence interval
Variable: median (yrs BP)
Description: Median modeled age of each depth interval in yrs BP
Variable: mean (yrs BP)
Description: Mean modeled age of each depth interval in yrs BP
Variable: Age CE (yrs BP)
Description: Modeled age of each depth interval in Common Era (CE) notation, calculated as 1950-mean age.
Sharing/Access information
Monthly precipitation, temperature, vapor pressure, and hydrogen isotope data were downloaded from the Global Network of Isotopes in Precipitation (GNIP) Water Isotope System for Electronic Retrieval (WISER). Water isotope data from GNIP can be found at the following address: https://www.iaea.org/services/networks/gnip
Please note that Rapa Nui is labeled as "Isla de Pascua" in the GNIP database. The Rapa Nui dataset is currently undergoing recompilation (7/2025), but should be reposted to the GNIP site shortly.
The age model for KAO08-03 was provided upon request by the authors of Rull et al., 2018.
Rull, V. et al. CLAFS, a Holistic Climatic-Ecological-Anthropogenic Hypothesis on Easter Island’s Deforestation and Cultural Change: Proposals and Testing Prospects. Front. Ecol. Evol. 6, (2018).
