Greater adaptation potential to climate change in populations of Quercus macrocarpa closer to edges of latitudinal gradient than those from mid-latitudes
Data files
Jan 23, 2026 version files 665.13 KB
-
heritability.zip
564.70 KB
-
README.md
7.30 KB
-
selection.zip
93.13 KB
Abstract
With climate change ongoing, tree populations will encounter variable selection pressures that may lead to local extinction if they are unable to adapt or respond appropriately. We use a reciprocal transplant experiment with the widespread tree species, Quercus macrocarpa L., across a latitudinal gradient to predict responses to selection expected with climate change. We asked 1) Is there genetic variation within populations for traits relevant to climate change? 2) Are there differences in direction and strength of selection for these traits in each garden? and 3) Do northern populations have evolutionary potential to adapt to selection in warmer, southern gardens? To address these questions, we estimated genetic variance for three sets of three traits grouped by physiology, spectral wavelengths, and morphology. We then conducted selection analysis to estimate directional and stabilizing selection on each of the traits in each garden and for each population. We used the Breeder’s equation to estimate the response to selection for each trait to assess their evolutionary potential. Our results indicate that traits related to morphology and growth are under strong directional selection in warmer gardens. We also found that populations closer to the edges of the species’ range have a high potential to adapt to climate change, due to their responses to selection in the warmest garden. We found low genetic variance in the population in the middle of the range, which may be a result of lower environmental stress. These results can help inform strategies to improve species resilience in the face of climate change as they tell us which populations are likely to be able to adapt to climate change, so managers planning assisted migration can source seeds from populations likely to adapt to climate change.
https://doi.org/10.5061/dryad.xwdbrv1ps
Description of the data and file structure
We measured physiology and morphology data in 3 bur oak common gardens in the growing season in 2023. We estimated genetic variance for three sets of three traits grouped by physiology, spectral wavelengths, and morphology. We then conducted selection analysis to estimate directional and stabilizing selection on each of the traits in each garden and for each population. We used the Breeder’s equation to estimate the response to selection for each trait to assess their evolutionary potential.
Files and variables
File: heritability.zip
Description: Data and code necessary to estimate heritability using QUERCUS software.
Description of files in heritability.zip
File name: h2 results.csv
Description: Genetic variance results of QUERCUS analysis extracted from QUERCUS program output
Column headers:
garden: common garden location (MN = Minnesota, IL=Illinois, OK = Oklahoma)
population: individual's population based on state of origin
trait: trait that data corresponds to (LMA = leaf mass per area, RGR = relative growth rate, thickness = leaf thickness, WBI = water band index, ANT = anthocyanin content, CCI = chlorophyll:carotenoid index, 700 = reflectance at 700 nm, 900 = reflectance at 900 nm, 1400 = reflectance at 1400 nm
analysis: grouping that traits were analysed in
Va: genetic variance
Ve: environmental variance
Log-like_Va:log likelihood value for genetic variance
Log-like_Va constrained: log likelihood of genetic variance with components constrained to 0
notes: notes on what QUERCUS Analysis file output the data was from
h2_full: broad sense heritability
LRT:likelihood ratio test value
df: degrees of freedom
Chi_sq dist: chi squared distribution
Significant?: Yes indicates significant at the p<0.05 level based on likelihood ratio test
Va_varcov: variance covariance for genetic variance
Va_SE: standard error for genetic variance
Ve_varcov: variance covariance for environmental variance
Ve_SE: standard error for environmental variance
File name: adaptation potential.csv
Description: Results of Breeder’s equation and data used to calculate
Column headers:
garden: common garden location (MN = Minnesota, IL=Illinois, OK = Oklahoma)
population: individual's population based on state of origin
trait: trait that data corresponds to (LMA = leaf mass per area, RGR = relative growth rate, thickness = leaf thickness, WBI = water band index, ANT = anthocyanin content, CCI = chlorophyll:carotenoid index, 700 = reflectance at 700 nm, 900 = reflectance at 900 nm, 1400 = reflectance at 1400 nm
S: Selection differential
h2_full: broad sense heritability
R(breeder's eq): Breeder's equation calculated from S and h2_full
File name: Heritability_post quercus analysis_clean.R
Description: R script to analyze data in h2 results.csv and adaptation potential.csv
Folder: quercus_program_files
Description: This folder contains folders for each population in each garden labeled as “garden.population” (for example il.ok contains data for the Oklahoma population in the Illinois garden). Each of the “garden.population” folders contains a .pas file that is run with a Pascal compiler in command line of your operating system using the command “fpc” followed by your file path. The .pas file can be edited in a text editing software. In addition, there are three folders (physio, spectra, and morpho) which contain sibships.txt files with data for each of the three traits as well as the parental relationships for running with the QUERCUS program, using the code in the .pas files. At the top of each sibship file is encoded information to give instructions to the QUERCUS program and tell it which components to constrain based on the covariance matrix. (More information about running QUERCUS can be found in the QUERCUS documentation here: https://cbs.umn.edu/eeb/about-eeb/helpful-links/quercus-quantitative-genetics-software). The folders also contain .csv sibships files with the data used to create the sibships.txt files. The column names for the sibship csv files contain the data for garden individuals IDs (individual_ID), maternal IDs (maternal_ID), paternal_IDs, the block the individual is located in in the garden (block), and measurements for 3 traits (which varies depending on the physio, spectral, or morpho analysis). Data that is missing for an individual is encoded with -99, as specified in the QUERCUS program. I have also filed the output files (which start with Analysis) into their respective folders.
File: selection.zip
Description: Data and code necessary to calculate selection gradients for physiological traits and morphological traits using R software.
Description of files in selection.zip
File name: spec 2023_resamp 100.csv
File description: resampled spectral reflectance data
Column headers:
garden_ID: unique identifier for individuals as garden.block.row.column
block: block in common garden
garden: garden location
mother_id: identifier for maternal tree that seed was collected from
700: 700 nm
900: 900 nm
1400: 1400 nm
File name: spectra_indices2023.csv
File description: spectral indices calculated from spectral reflectance data
garden_ID: unique identifier for individuals as garden.block.row.column
CCI: chlorophyll:carotenoid index
WBI: Water band index
File name: leaf thickness.csv
File description: leaf thickness data measured using calipers
garden_ID: unique identifier for individuals as garden.block.row.column
leaf_thickness_mm: leaf thickness in millimeters. Blank values indicate where individuals were not measured due to lack of sufficient leaf material.
File name: prospect_traits.csv
File description: traits modeled from spectral reflectance data using the PROSPECT models
garden_ID: unique identifier for individuals as garden.block.row.column
block: block in common garden
garden: garden location
population: population that individual originated from
mother_id: identifier for maternal tree that seed was collected from
ANT: anthocyanins
LMA: leaf mass per area
File name: growth calcs 2023.1.csv
File description: relative growth rate from 2021-2023
garden_ID: unique identifier for individuals as garden.block.row.column
mother_id: identifier for maternal tree that seed was collected from
garden: garden location
population: population that individual originated from
block: block in common garden
garden_block: identifier for block in garden
RGR_Vol: relative growth rate for stem volume, in mm^3/day. NA indicates dead individuals that do not have growth status.
deadmissing_fall23: survival status of individuals when surveyed in 2023, with 1 indicating alive and 0 indicating dead individuals.
Code/software
QUERCUS software: https://cbs.umn.edu/eeb/about-eeb/helpful-links/quercus-quantitative-genetics-software
R software v.4.4.1 using packages: aster, gdata, MASS, car, lme4, lattice, grDevices, scatterplot3d, tidyverse
