Stomatal conductance and tree growth response to multi-year droughts in fire-maintained and fire-excluded forests
Data files
Jun 23, 2025 version files 4.06 GB
-
All_2022_Cores_Q2_6-6-24_Tucson.TXT
376.56 KB
-
Atmospheric_CO2.csv
4.03 KB
-
Fire_Occurences_Corrected_1-15-23.xlsx
18.23 KB
-
Fire_recurrence_shapefile.cpg
5 B
-
Fire_recurrence_shapefile.dbf
89.71 MB
-
Fire_recurrence_shapefile.prj
425 B
-
Fire_recurrence_shapefile.sbn
109.28 KB
-
Fire_recurrence_shapefile.sbx
3.12 KB
-
Fire_recurrence_shapefile.shp
7.65 MB
-
Fire_recurrence_shapefile.shp.xml
80.42 KB
-
Fire_recurrence_shapefile.shx
91.64 KB
-
Gila_DEM.tfw
83 B
-
Gila_DEM.tif
2.81 GB
-
Gila_DEM.tif.aux.xml
2.83 KB
-
Gila_DEM.tif.ovr
949.80 MB
-
Gila_hillshade.tfw
90 B
-
Gila_hillshade.tif
144.76 MB
-
Gila_hillshade.tif.aux.xml
2.56 KB
-
Gila_hillshade.tif.ovr
50.07 MB
-
Gila_hillshade.tif.vat.cpg
5 B
-
Gila_hillshade.tif.vat.dbf
7.75 KB
-
Gravin_2017_corrections.csv
5.31 KB
-
Growth_Competition_drought_data_5-26-23.xlsx
40.84 KB
-
isotope_growth_key.csv
179.62 KB
-
Job_995_Willson_Cellulose_C_Summary_Trays_1-5.xlsx
33.51 KB
-
Job_995_Willson_Cellulose_C_Summary_Trays_11-22.xlsx
74.10 KB
-
Job_995_Willson_Cellulose_C_Summary_Trays_6-10.xlsx
39.43 KB
-
Nested_Data_withQ1_Data_5-23-22.xlsx
114.23 KB
-
PRISM_Monthly_Climate_1895-1900_SPEI.csv
300.73 KB
-
PRISM_Monthly_Climate_1900-1915_SPEI.csv
899.76 KB
-
PRISM_Monthly_Climate_1915-1930_SPEI.csv
900.20 KB
-
PRISM_Monthly_Climate_1930-1945_SPEI.csv
900.33 KB
-
PRISM_Monthly_Climate_1945-1960_SPEI.csv
899.70 KB
-
PRISM_Monthly_Climate_1960-1975_SPEI.csv
899.62 KB
-
PRISM_Monthly_Climate_1975-1990_SPEI.csv
901.03 KB
-
PRISM_Monthly_Climate_1990-2005_SPEI.csv
900.80 KB
-
PRISM_Monthly_Climate_2005-2020_SPEI.csv
900.72 KB
-
README.md
22.35 KB
Abstract
In the western US, increased tree density in dry conifer forests from fire exclusion has caused tree growth declines, which is being compounded by hotter multi-year droughts. The reintroduction of frequent, low-severity wildfires reduces forest density by removing fire-intolerant trees, which can reduce competition for water and improve tree growth response to drought. We assessed how lower forest density following frequent, low-severity wildfire affected tree stomatal conductance and growth response to drought by coring and measuring competition surrounding ponderosa pines (Pinus ponderosa) in the Gila and Apache-Sitgreaves National Forests, NM, USA, that either experienced 3-5 fires following long-term fire-exclusion or remained fire-suppressed. We quantified tree growth decline to (resistance) and how quickly growth recovered (resilience) from drought to two recent multi-year droughts and compared values between trees in fire-maintained and fire-excluded forests. We assessed stomatal conductance among trees by sampling wood from tree rings to measure stable carbon isotopes during and after both droughts, which we used to calculate evaporative water use efficiency. Trees in fire-maintained forests had greater resistance than trees in fire-excluded forests during the first drought, but growth responses became similar once the first drought ended. Interestingly, growth responses rarely varied despite evaporative water use efficiency increasing two times faster among trees in fire-excluded forests after the first drought commenced. Post-drought growth responses primarily varied by aspect, with trees on northerly aspects exhibiting greater resilience to both droughts than trees on southerly aspects. Our results indicate that while trees had density-independent growth responses to drought, trees in fire-maintained forests were less water stressed than trees in fire-excluded forests. Therefore, the reintroduction of frequent, low-severity wildfire regimes has the potential to moderate some effects of hotter droughts as climate change intensifies.
https://doi.org/10.5061/dryad.d7wm37q9w
Description of the data and file structure
Data and Code for "Trees in fire-maintained forests have similar growth responses to drought, but greater stomatal conductance than trees in fire-excluded forests"
Created by Kevin G. Willson
Files and variables
File: All_2022_Cores_Q2_6-6-24_Tucson.TXT
Description: This textfile contains growth data for all tree cores collected for analysis. Columns are as follows: First column = tree core series identification, Second column = decade affiliated with growth measurements, Third-Twelth columns = tree-ring width measurements (in thousands of mm).
File: Atmospheric_CO2.csv
Description: This dataframe contains annual average concentrations of atmospheric CO2 used to calculate eWUE in the analysis. Columns are as follows: "year" = year of average CO2 values, "mean" = mean concentration of atmospheric CO2, "unc" = estimated uncertainty in the mean (standard deviation).
File: Fire_Occurences_Corrected_1-15-23.xlsx
Description: This dataframe contains number of fires and years of fires that burned through each plot, which was aggregated from monitoring trends in fire severity data, corrected fire boundaries by Parks et al. 2015, and historical fire boundaries from Rollins et al. 2002. Columns are as follows: "Fire", "Patch", and "Transect" = identification for each plot, "Num_fires" = the number of fires each plot experienced, "Fire_year_1" = the year a plot experienced its first fire, "Fire_year_2" = the year a plot experienced its second fire, "Fire_year_3" = the year a plot experienced its third fire, "Fire_year_4" = the year a plot experienced its fourth fire, "Fire_year_5" = the year a plot experienced its fifth fire.
File: Fire_recurrence_shapefile.sbx
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Fire_recurrence_shapefile.sbn
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Fire_recurrence_shapefile.cpg
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Gravin_2017_corrections.csv
Description: This dataframe contains records of carbon isotopes in atmospheric CO2. Columns are as follows: "Date" = year of value, "NH Delta14co2" = annual mean of d14co2 in the northern hemisphere, "Tropics Delta14co2" = annual mean of d14co2 in the tropics, "SH Delta14co2" = annual mean of d14co2 in the southern hemisphere, "Global delta13co2" = annual mean of d13co2 around the world.
File: Growth_Competition_drought_data_5-26-23.xlsx
Description: This dataframe contains stand structure and competition data for each tree with stable carbon isotope measurements used in this project. Columns are as follows: "Fire" = The number of fires each tree experienced, "Patch" = the delineation of expected hillslope for each sampled plot, "Transect" = the delineation of expected aspect for each sampled plot, "Tree" = The number of each cored tree in each plot, "Comp_Density" = the density (/ha) of trees within 5 m of the tree, "Comp_BA" = the total basal area (/ha) of trees within 5 m of the tree, "Comp_QMD" = the average size of trees (dbh in cm) within 5 m of the tree, "NestedDensity" = stand density (/ha) of the plot, "Nested_OverstoryPIPODensity" = stand density (/ha) of ponderosa pine trees > 25 dbh, "Nested_OverstoryDensity" = stand density (/ha) of trees > 25 dbh, "Nested_PIPO_Density" = stand density (/ha) of ponderosa pine trees, "NestedBA" = stand basal area (/ha), "Nested_PIPO_BA" = stand basal area (/ha) of ponderosa pine, "Nested_QMD" = average tree size (dbh in cm) in the plot, "Nested_PIPO_QMD" = average ponderosa pine tree size (dbh in cm) in the plot, "Slope" = average hillslope (degrees) of the plot, "CTI" = topographic wetness index of the plot, "Elevation" = plot elevation (m), "Aspect" = plot aspect (degrees), "age" = tree age at coring, "DBH" = tree diameter at breast height (cm) at time of coring, "area" = growing space (sq m) around each tree as measured using voronoi diagrams, "heygi_sum" = Heygi index for each tree, "RK_sum" = Rouvinen and Kuuluvainen measure of competition.
File: Gila_hillshade.tfw
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: isotope_growth_key.csv
Description: This dataframe contains growth and intrinsic water use efficiency values aligned by tree and year to help merge dataframes together. Columns are as follows: "Fire" = The number of fires each tree experienced, "Patch" = the delineation of expected hillslope for each sampled plot, "Transect" = the delineation of expected aspect for each sampled plot, "Tree" = The number of each cored tree in each plot, "iWUE" = intrinsic water use efficiency values calculated using carbon isotope data, "fire_effect_pos" = Delineation of years that fire-maintained trees experienced fire and separation between trees from fire-maintained and fire-excluded forests, "Detrended_growth" = tree-ring width index values (unitless) for each tree by year collected from tree cores and detrended using the modified negative exponential detrending methods, "Year" = year of data point, "Fire_occurred" = Delineation of trees from fire-maintained and fire-excluded forests.
File: Fire_recurrence_shapefile.shp
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Job_995_Willson_Cellulose_C_Summary_Trays_11-22.xlsx
Description: 1960 isotope results that represent 2162 years of stable carbon isotope discrimination values from 1996-2018 of 94 trees. Column "Identifier 1" contains a unique tree and year identifier. Column "%C" represents percent carbon in the sample. Column "d 13C/12C" represents the uncorrected carbon discrimination value of the sample. Column "d13C corr" represents corrected carbon discriminiation value of the sample used to calculate intrinsic water use efficiency.
File: Fire_recurrence_shapefile.shx
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Gila_hillshade.tif.vat.cpg
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: Gila_DEM.tif.aux.xml
Description: 10 meter elevation raster across the Gila national forest used to calculate hillshade in the Gila wilderness.
File: PRISM_Monthly_Climate_1900-1915_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1900,-1915 which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: PRISM_Monthly_Climate_1915-1930_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1915-1930, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: PRISM_Monthly_Climate_1930-1945_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1930-1945, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: PRISM_Monthly_Climate_1960-1975_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1960-1975, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: PRISM_Monthly_Climate_2005-2020_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 2005-2020, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: Gila_DEM.tfw
Description: 10 meter elevation raster across the Gila national forest used to calculate hillshade in the Gila wilderness.
File: PRISM_Monthly_Climate_1975-1990_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1975-1990, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: Fire_recurrence_shapefile.prj
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Fire_recurrence_shapefile.shp.xml
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: PRISM_Monthly_Climate_1895-1900_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1895-1900, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: Gila_hillshade.tif.aux.xml
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: Gila_hillshade.tif.vat.dbf
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: Job_995_Willson_Cellulose_C_Summary_Trays_1-5.xlsx
Description: 1960 isotope results that represent 2162 years of stable carbon isotope discrimination values from 1996-2018 of 94 trees. Column "Identifier 1" contains a unique tree and year identifier. Column "%C" represents percent carbon in the sample. Column "d 13C/12C" represents the uncorrected carbon discrimination value of the sample. Column "d13C corr" represents corrected carbon discriminiation value of the sample used to calculate intrinsic water use efficiency.
File: Job_995_Willson_Cellulose_C_Summary_Trays_6-10.xlsx
Description: 1960 isotope results that represent 2162 years of stable carbon isotope discrimination values from 1996-2018 of 94 trees. Column "Identifier 1" contains a unique tree and year identifier. Column "%C" represents percent carbon in the sample. Column "d 13C/12C" represents the uncorrected carbon discrimination value of the sample. Column "d13C corr" represents corrected carbon discriminiation value of the sample used to calculate intrinsic water use efficiency.
File: PRISM_Monthly_Climate_1945-1960_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1945-1960, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: PRISM_Monthly_Climate_1990-2005_SPEI.csv
Description: This dataframe contains monthly climate data for all plots in the study from 1990-2005, which was used to quantify annual VPD values for that period and to assess seasonal climate-growth correlation patterns with temperature and precipitation. Columns are as follows: "Name" = identification for each plot, "Longitude" = longitude for each plot location, "Latitude" = latitude for each plot location, "Elevation" = elevation (in meters) of the centroid for the 4-km grid cell the plot occurred in, "Date" = Year and month of the climate data point, "ppt (mm)" = total precipitation for that year/month in millimeters, "tmin (degrees C)" = minimum temperature recorded during that year/month in degrees Celsius, "tmax (degree C)" = maximum temperature recorded during that year/month in degrees Celsius, "vpdmax (hPa)" = maximum vapor pressure deficit recorded during that year/month in Hectopascals.
File: Nested_Data_withQ1_Data_5-23-22.csv
Description: This dataframe contains stand structure data for the 42 plots included in the study. Columns are as follows: "Date" = The dat when data was collected, "Week" = the week number, representing which trip in which the data was collected, "Question" = the dissertation question in which the data was collected for, "Fire" = The number of fires each tree experienced, "Patch" = the delineation of expected hillslope for each sampled plot, "Transect" = the delineation of expected aspect for each sampled plot, "Species" = the species of an individual tree, "DBH" = diameter at breast height, measured in cm, "Tree_Num" = cored trees in each plot delineated by number for sample identification, "Height" = the height of some trees measured in m.
File: Fire_recurrence_shapefile.dbf
Description: Shapefile of fire history in the Gila from 1909-2018 used to make Figure 1. Locations did not experience any fire until 1950 at the earliest.
File: Gila_hillshade.tif.ovr
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: Gila_hillshade.tif
Description: 10 meter hillshade of the Gila wilderness used to make Figure 1.
File: Gila_DEM.tif.ovr
Description: 10 meter elevation raster across the Gila national forest used to calculate hillshade in the Gila wilderness.
File: Gila_DEM.tif
Description: 10 meter elevation raster across the Gila national forest used to calculate hillshade in the Gila wilderness.
File: Supplemental_figures.doc
Description: Seven supplemental figures that: 1) compare tree age and tree size between forests, 2) SPEI over the study period, 3) evaporative and 4) intrinsic water use efficiency against growing space, tree size, and growth rates, 5) vapor pressure deficit over the study period, 6) differences in competition between forests, and 7) isotopic trends over the study period.
File: Supplemental_Tables_S1-S7.doc
Description: Seven supplemental tables that show: 1-4) outputs of linear mixed models with ΔAIC scores < 5 for resistance and resilience to the first and second droughts, 5) outputs of linear mixed models with ΔAIC scores < 5 for evaporative water use efficiency, 6) descriptive statistics for non-climate predictor variables included in model selection, and 7) descriptive statistics for climate predictor variables included in model selection.
Code/software
Code:
1 - Question2-3_Averaging_Cores_together.R = Code to remove undateable cores and average growth between series for each tree.
2 - Detrending_tree_growth_code.R = Code to detrend growth data among trees from fire-maintained forests and create a long dataframe to use in future analyses.
3 - Analysis pt 1 - 4 or 7 -year drought comparisons.R = Code to assesses differences in growth response between trees in fire-maintained (FM) and fire-excluded forests (FE) during and after two droughts from 2000-2014 droughts.
4 - Analysis pt 2 - candidate predictor variable organization by fm and fe.R = Code to aggregate climate, competition, fire history, topography, and tree characteristic variables to explain tree growth response to the two droughts.
5 - Analysis pt 2.5 - calculating leaf area (and abg) for ewue regression.R = Code to aggregate additional predictor variables we used to explain annual evaporative water use efficiency among trees.
6 - Analysis pt 3 - candidate predictor variable selection.R = Code to run correlation analysis that selected candidate predictor variables to explain tree growth response to the two droughts.
7 - Analysis pt 3.5 - climwin = Code to run climwin analysis, which selected specific climate windows that best explained tree growth response and eWUE during and after the two droughts.
8 - Analysis pt 4 - growth, resistance, resilience analysis.R = Code to run model selection with AICc values to determine the best model for explaining tree growth reponses to the two droughts.
9 - Analysis pt 5 - isotopic comparison.R = Code to make figures that assess growth and water use for trees with annual isotopic data by fire-maintained and fire excluded trees.
10 - Analysis pt 6 - annual ewue regression.R = Code to aggregate potential predictor variables, run correlation analysis, and run model selection to determine the best model for explaining annual evaporative water use efficiency values among trees.
11 - Supplemental tables S6 and S7 creation.R = Code for quantifying means and ranges of predictor variables used to create supplemental tables S6 and S7.
