Data from: A decision support tool using an open-source methodology for identifying Woody encroachment and Juniper species vulnerability in the Chickasaw Nation, Oklahoma, USA
Data files
Jan 16, 2026 version files 5.38 GB
-
Ecology_Evolution_Manuscript_Files.zip
271.21 KB
-
GEE_Model_Files.zip
5.38 GB
-
README.md
22.21 KB
Abstract
As a result of 20th century farming practices, fire suppression, grazing pressures, federal agriculture policies, drought, and related factors, juniper trees have experienced a dramatic increase in abundance across the Great Plains region, including millions of acres in Oklahoma. This invasive vegetation poses a significant threat to the environment, soil health, cultural resources, and economies within the Cross Timbers region of the Chickasaw Nation in south-central Oklahoma. To combat these negative impacts, the Chickasaw Nation (CN) has undertaken a proactive approach to evaluate the extent of juniper encroachment within the CN treaty territory (CNTT) and to use this information to make an economic case to landowners for the implementation of both prescribed burns and juniper removal in general. This article describes the creation of a species-focused evaluation methodology, model, and decision-support tool, called the Juniper Evaluation Tool (JET), that is simple, open-source, and applicable at an appropriate spatial scale to assist landowners, land managers, and CN resource managers in targeting best management practices and slow the encroachment of juniper tree species. The JET, a product of a Natural Resources Conservation Service (NRCS) Conservation Innovation Grant (CIG), builds on the methods and functions of other planning tools, such as the Rangeland Analysis Platform (RAP) and the Rangeland Brush Estimation Toolbox (RaBET), but the outputs are tailored to parcel level management for watershed conservation planning in the CNTT. The JET aims to be a resource that is replicable, scalable, and actionable, and may provide the basis for an expanded tool that can be applied in other parts of Oklahoma and the Great Plains.
Dataset DOI: 10.5061/dryad.31zcrjf1s
Description of the data and file structure
A. Download the Ecology_Evolution_Manuscript_Files.zip folder and extract all files, which include one folder and 24 individual files.
The following is a description of all the files extracted from the Ecology_Evolution_Manuscript_Files.zip folder:
FOLDER
1. FINAL Validation Data
This folder contains the shapefile (and associated files) of all the validation data used for the model accuracy assessment. It can be opened in QGIS, ArcGIS Pro or any other spatial analysis program that can handle shapefiles. A Universal Transverse Mercator projection in zone 14 and the 1983 North American Datum was used for the shapefile.
The landcover types for the 1, 443 validation points in the shapefile are as follows:
0 -- Non-Juniper Tree (bare ground, unimproved roadway, pasture, crop land, deciduous tree species, etc.)
3 -- Edge Juniper Trees (an individual juniper tree in a linear pattern, usually along a fence, roadway or edge of a property boundary)
4 -- Lone Juniper Tree (an isolated individual juniper tree)
5 -- Stand of Junipers (more than 6 juniper trees near or next to each other)
6 -- Group of Junipers (less than 6 junipers trees near or next to each other)
FILES
1. County Encroachment Level_JET Data Export.xlsx
This spreadsheet contains one tab (County Level_JET Download) that has data that was exported from the Juniper Evaluation Tool (JET) and lists the amount of juniper species by county. Rows include the 13 counties represented in the study area. Columns A - G include:
County Name – the name of each county represented in the study area.
County Area (Acres) – the area of each county in acres.
Tree Cover Area (Acres) - the distribution of total tree cover in acres by county (calculated directly from the juniper tree species distribution model output and includes woodland transition acres).
Tree Cover (%) - the percentage of tree cover by county (calculated as a percent between tree cover and county acres).
Woodland Transition (Acres) - the distribution of juniper species in acres by county (calculated directly from the juniper tree species distribution model and a subset of total tree cover acres).
Woodland Transition (%) - the percentage of the distribution of juniper species by county (calculated as a percent between the woodland transition acres and county acres).
Percent Juniper Cover with Respect to Tree Cover - the percentage of juniper cover found amongst the tree cover by county (i.e., what percent of the tree cover is made up of juniper species) (calculated as a percent between woodland transition area and tree cover area).
2. County Encroachment Level_JET Data Export.csv
This spreadsheet contains one tab (County Encroachment Level_JET D) and is a *.csv version of County Encroachment Level_JET Data Export.xlsx.
3. Watershed Encroachment Level_JET Data Export.xlsx
This spreadsheet contains one tab (Watershed Level_JET Download) that has data that was exported from the Juniper Evaluation Tool (JET) and lists the amount of juniper species by watershed. Rows include the 41 watersheds represented in the study area. The first table (rows 1-42) is an alphabetized list and the second table (rows 43-84) is the same table sorted by column G.
Columns A – G include:
Watershed Name - the name of each watershed represented in the study area.
Watershed Area (Acres) – the area of each watershed in acres.
Tree Cover Area (Acres) - the distribution of total tree cover in acres by watershed (calculated directly from the juniper tree species distribution model output and includes woodland transition acres).
Tree Cover (%) - the percentage of tree cover by watershed (calculated as a percent between tree cover and watershed acres).
Woodland Transition (Acres) - the distribution of juniper species in acres by watershed (calculated directly from the juniper tree species distribution model and a subset of total tree cover acres).
Woodland Transition (%) - the percentage of the distribution of juniper species by watershed (calculated as a percent between the woodland transition acres and watershed acres).
Percent Juniper Cover with Respect to Tree Cover - the percentage of juniper cover found amongst the tree cover by watershed (i.e., what percent of the tree cover is made up of juniper species) (calculated as a percent between woodland transition area and tree cover area).
4. Watershed Encroachment Level_JET Data Export.csv
This spreadsheet contains one tab (Watershed Encroachment Level_JE) and is a *.csv version of Watershed Encroachment Level_JET Data Export.xlsx.
5. GEE CODE LINK.docx
This document contains the link to the google earth engine (GEE) code used to create the model for the JET.
6. GEE CODE LINK.pdf
This document contains the link to the google earth engine (GEE) code used to create the model for the JET.
7. NRCS_CIG_Accuracy_Assessment.xlsx
This spreadsheet contains one tab (All Field & Aerial Data) of validation data used for the accuracy assessment with calculations provided.
A. Data $A$:$C$1:1443 contains 1,443 rows of validation data and corresponds to the files in the FINAL Validation Data.zip folder. The three columns include:
· Land Cover Code
0 -- Non-Juniper Tree (bare ground, unimproved roadway, pasture, crop land, deciduous tree species, etc.)
3 -- Edge Juniper Trees (an individual juniper tree forming a linear pattern, usually along a fence, roadway or edge of a property boundary)
4 -- Lone Juniper Tree (an isolated individual juniper tree)
5 -- Stand of Junipers (more than 6 juniper trees near or next to each other)
6 -- Group of Junipers (less than 6 junipers trees near or next to each other)
· Land Cover Type
Non- Cedar = Non-Juniper Tree (bare ground, unimproved roadway, pasture, crop land, deciduous tree species, etc.)
Edge = Edge Juniper Trees (an individual juniper tree forming a linear pattern, usually
along a fence, roadway or edge of a property boundary)
Lone = Lone Juniper Tree (an isolated individual juniper tree)
Stand = Stand of Junipers (more than 6 juniper trees near or next to each other)
Group = Group of Junipers (less than 6 junipers trees near or next to each other)
· Sample_1 (0= non juniper 1 = juniper)
The woodland transition area data from the model output in the Juniper Evaluation Tool (JET) was intersected with the 1,443-validation land cover type points to produce this column containing zeros and ones, representing non juniper and juniper output, respectively. This column was used in the accuracy assessment tables (Data $E$:$T$55).
B. Data $E$:$K$3:6 (Validation Data Table)
This table was generated using the data ($A$:$C$1:1443) in the table, specifically columns B and C.
C. Data $N$:$T$2:45 (Accuracy Assessment Raw Data Table)
This table was generated using the data in the Validation Data Table ($E$:$K$3:6). The reference data was entered in by hand from the Validation Data Table and User’s Accuracy (UA) and Producer’s Accuracy (PA) values calculated with formulas (provided in cell). Overall Accuracy (OvAc) and the Kappa Coefficient (KHAT) were also calculated with formulas (provided in each cell). Lone, Edge, Group, Stand and Combined assessment data was calculated.
D. Data $E$:$I$13:19 (Producer’s and User’s Accuracy Table)
This table was generated from the cells in the Accuracy Assessment Raw Data Table ($N$:$T$2:45).
E. Data $E$:$F$23:26 (Grouped Overall Accuracy Table)
This table was generated from the cells in the Validation Data Table ($E$:$K$3:6). It shows the combined results of each point class and removing one until the stand point class remains.
F. Data $H$:$J$22:27 (Overall Accuracy and Kappa Coefficient by Class Table)
This table was generated from the cells in the Accuracy Assessment Raw Data Table ($N$:$T$2:45).
G. Data $E$:$L$31:36 (Manuscript Data, Table 7)
This table is a cleaned up version of the Validation Data Table (B. Data $E$:$K$3:6) for the manuscript.
H. Data $E$:$K$40:46 (Manuscript Data, Table 8)
This table is a cleaned up combined version of the Producer’s and User’s Accuracy Table (D. Data $E$:$I$13:19) and the Overall Accuracy and Kappa Coefficient by Class Table (F. Data $H$:$J$22:27) for the manuscript.
I. Data $E$:$F$51:55 (Manuscript Data, Table 9)
This table is a cleaned up version of the Grouped Overall Accuracy Table (E. Data $E$:$F$23:26 for the manuscript.
8. NRCS_CIG_Accuracy_Assessment.csv
This spreadsheet contains one tab (NRCS_CIG_Accuracy_Assessment) and is a *.csv version of NRCS_CIG_Accuracy_Assessment.xlsx.
9. RAP vs JET Tree Cover Time Series Statistics.xlsx
This spreadsheet contains 15 tabs of data downloaded from the Rangeland Analysis Platform (RAP) 30-meter vegetation cover data (https://rangelands.app/). This data was used to compare the RAP tool with the JET (e.g. Comparison tab). The yellow highlight in the TRE column for each county tab shows percent tree cover for 2021-22 and averaged to produce the value further down in the column. These data were then compiled the Comparison tab for analysis.
Each individual tab is explained in further detail below:
A. Bryan County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
B. Carter County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
C. Coal County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
D. Garvin County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
E. Grady County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
F. Jefferson County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
G. Johnston County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022. Two sets of data are shown for comparison.
H. Love County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
I. Marshall County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
J. McClain County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
K. Murray County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
L. Pontotoc County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
M. Stephens County
This tab contains the entire county data downloaded from the RAP tool. The TRE column represents percent tree cover. Because only part of the county is in the study area, the county was clipped to represent the portion within the study area and then downloaded. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET since this matches the two years used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
N. CN Data
This tab contains the data for the entire study area downloaded from the RAP tool. The TRE column represents percent tree cover. The highlighted data in yellow that represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The number at the bottom of the TRE column represents the average of both years, 2021 and 2022.
O. Comparison
This tab contains all the TRE column (i.e. percent tree cover) data copied from the previous tabs represented by year and county, rows and columns, respectively. The highlighted data in yellow represents data from 2021 and 2022 and was used for comparison to the JET as this data matches the data used in the model. The numbers at the bottom of each column represents the average of both years, 2021 and 2022. These averages were used to compare the tree cover data between the RAP tool and the JET. A summary of this comparison is represented in various ways in the following data block ($R$:$AF$1:56). The table (Table 2) was used in the supplemental material document. The t-Stat and p-value from the paired two sample for means t-test was used in the manuscript.
10. RAP vs JET Tree Cover Time Series Statistics_BryanCountySheet.csv
This spreadsheet is a *.csv version of the Bryan County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
11. RAP vs JET Tree Cover Time Series Statistics_CarterCountySheet.csv
This spreadsheet is a *.csv version of the Carter County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
12. RAP vs JET Tree Cover Time Series Statistics_CoalCountySheet.csv
This spreadsheet is a *.csv version of the Coal County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
13. RAP vs JET Tree Cover Time Series Statistics_GarvinCountySheet.csv
This spreadsheet is a *.csv version of the Garvin County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
14. RAP vs JET Tree Cover Time Series Statistics_GradyCountySheet.csv
This spreadsheet is a *.csv version of the Grady County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
15. RAP vs JET Tree Cover Time Series Statistics_JeffersonCountySheet.csv
This spreadsheet is a *.csv version of the Jefferson County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
16. RAP vs JET Tree Cover Time Series Statistics_JohnstonCountySheet.csv
This spreadsheet is a *.csv version of the Johnston County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
17. RAP vs JET Tree Cover Time Series Statistics_LoveCountySheet.csv
This spreadsheet is a *.csv version of the Love County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
18. RAP vs JET Tree Cover Time Series Statistics_MarshallCountySheet.csv
This spreadsheet is a *.csv version of the Marshall County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
19. RAP vs JET Tree Cover Time Series Statistics_McClainCountySheet.csv
This spreadsheet is a *.csv version of the McClain County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
20. RAP vs JET Tree Cover Time Series Statistics_MurrayCountySheet.csv
This spreadsheet is a *.csv version of the Murray County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
21. RAP vs JET Tree Cover Time Series Statistics_PontotocCountySheet.csv
This spreadsheet is a *.csv version of the Pontotoc County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
22. RAP vs JET Tree Cover Time Series Statistics_StephensCountySheet.csv
This spreadsheet is a *.csv version of the Stephens County sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
23. RAP vs JET Tree Cover Time Series Statistics_CN DataSheet.csv
This spreadsheet is a *.csv version of the CN Data sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
24. RAP vs JET Tree Cover Time Series Statistics_ComparisonSheet.csv
This spreadsheet is a *.csv version of the Comparison sheet in the RAP vs JET Tree Cover Time Series Statistics.xlsx file.
B. Download the GEE_Model_Files.zip folder and extract all files which include five folders. These folders contain the spatial data that were uploaded to Google Earth Engine because the data was not available in their online data repository.
The following is a description of all the folders extracted from the GEE_Model_Files.zip folder:
FOLDERS
1. Canopy_Height_Model
This folder contains 21 LiDAR-derived *.tif images used in the model as discussed in the manuscript.
2. LULC
This folder contains two land use land cover *.tif images used in the model as discussed in the manuscript.
3. Municipal_Parcels
This folder contains the shapefile (and associated files) of municipal parcels used in the model as discussed in the manuscript.
4. Roadways
This folder contains the shapefile (and associated files) of roadways used in the model as discussed in the manuscript.
5. Study_Area
This folder contains the shapefile (and associated files) of the study area (i.e., the Chickasaw Nation Treaty Territory) used in the first step of the model as discussed in the manuscript.
C. Download the Figures.zip folder and extract the two Figurefolders. These folders contain all figures contained within the manuscript and the supplemental document.
The following is a description of the folders extracted from the Figures.zip folder:
1. Manuscript Figures
This folder contains all manuscript figures and tables in various file formats and sizes. Please ask the authors for permission before use. Standard naming convention is used (e.g. Figure 1, Figure 2, Table 1, Table 2, etc.).
2. Supplemental Figures
This folder contains all supplemental document figures and tables in various file formats and sizes. Please ask the authors for permission before use. Standard naming convention is used (e.g. Figure 1, Figure 2, Table 1, Table 2, etc.).
Files and variables
File: Ecology_Evolution_Manuscript_Files.zip
Description: Provided in the previous "Data description" section.
Code/software
Microsoft Word and Excel Viewer, Photo Viewer, QGIS or ArcGIS Pro, Browser to view Google Earth Engine Code
