Data from: Contemporary fires are less frequent but more severe in dry conifer forests of the southwestern United States
Data files
Sep 06, 2024 version files 89.46 KB
-
field_data_all_trees.csv
40.14 KB
-
field_data_seed.csv
4.25 KB
-
field_data_sev.csv
8.12 KB
-
field_data_site.csv
25.37 KB
-
field_data_treecount.csv
4.58 KB
-
README.md
7 KB
Sep 10, 2024 version files 89.42 KB
-
field_data_all_trees.csv
40.14 KB
-
field_data_seed.csv
4.25 KB
-
field_data_sev.csv
8.12 KB
-
field_data_site.csv
25.37 KB
-
field_data_treecount.csv
4.58 KB
-
README.md
6.96 KB
Abstract
Wildfires in the southwestern United States are increasingly frequent and severe, but whether these trends exceed historical norms remains contested. Here we combine dendroecological records, satellite-derived burn severity, and field measured tree mortality to compare historical (1700-1880) and contemporary (1985-2020) fire regimes at tree-ring fire-scar sites in Arizona and New Mexico. We found that contemporary fire frequency, including recent, record fire years, is still <20% of historical levels. Since 1985, the fire return interval averages 58.8 years, compared to 11.4 years before 1880. Fire severity, however, has increased. At sites where trees historically survived many fires over centuries, 42% of recent fires resulted in high tree mortality. Suppressed wildfires tended to burn more severely than prescribed burns and fire use wildfires. These findings suggest that expanded use of low-severity prescribed and managed fire would help restore forest resilience and historical fire regimes in southwestern dry conifer forests.
https://doi.org/10.5061/dryad.98sf7m0sn
Description of the data and file structure
Field data were collected in Survey123, then adapted into .csv files for import and analysis in R. Sampling was performed in June - August 2021, and involved visits to 74 previously sampled tree-ring fire-scar sites. At each site center or found sampled tree, a 10-meter radius plot was installed, in which we recorded tree diameter at breast height (dbh), species, and status (live or dead). We measured diameter and assigned species for downed logs and recorded an overall count of trees both live and dead, standing and down. We also completed a modified (simplified) CBI protocol to assess severity, and counted seedlings by species across the plot. Sampling was focused on 6 geographic areas where networks of fire history sites had been established prior to wildfires occurring over the past ten years (2011-2020): the Jemez, Rincon, Santa Catalina, Pinaleño, and Chiricahua Mountains, and the Kaibab Plateau.
Tree species codes explained
| ABCO | Abies concolor | White fir |
|---|---|---|
| JUDE | Juniperus deppeana | Alligator juniper |
| JUSC | Juniperus scopulorum | Rocky Mountain juniper |
| PIAZ | Pinus arizonica | Arizona pine |
| PIED | Pinus edulis | Piñon pine |
| PIEN | Picea engelmannii | Engelmann spruce |
| PIFL | Pinus flexilis | Limber pine |
| PIPO | Pinus ponderosa | Ponderosa pine |
| PISF | Pinus strobiformis | Southwestern white pine |
| PPTR | Populus tremuloides | Quaking aspen |
| PSME | Pseudotsuga menziesii | Douglas-fir |
| QUGA | Quercus gambellii | Gambel oak |
Files and variables
File: field_data_all_trees.csv
Description: all live and dead trees, standing and down, were identified to species and assessed as being killed by fire (if dead).
Variables
- Site Name: identification code for field plot
- ID: identification code for original fire history site
- species: tree species code (spelled out in metadata)
- DBH: diameter at breast height
- Status: live or dead at time of field sampling
- Killed by fire: qualitative assessment by field data collectors as to whether tree was killed by most recent fire
- NA values derived from original metadata (signify missing data)
File: field_data_seed.csv
Description: all seedlings were counted and identified to species in a 5-meter subplot within the larger plot area, centered at plot center
Variables
- Site Name: identification code for field plot
- ID: identification code for original fire history site
- mtnrange: geographic area of interest (one of six identified for field sampling)
- PIPO: plot seedling count for ponderosa pine
- PIST: plot seedling count for southwestern white pine
- PIAR5: plot seedling count for Arizona pine
- PSME: plot seedling count for Douglas fir
- ABCO: plot seedling count for white fir
- PIEN: plot seedling count for Engelmann spruce
- JUOS: plot seedling count for juniper
- POTR: plot seedling count for quaking aspen
- count_conifer: plot seedling count summed for all conifer species
- count_total: plot seedling count summed for all species
File: field_data_sev.csv
Description: a modified CBI (Composite Burn Index, typically conducted one-year post-fire) plot protocol was utilized to assess severity, with particular focus on trees and heavy fuels. all measurements performed by ocular estimation.
Variables
- Site Name: identification code for field plot
- ID: identification code for original fire history site
- mtnrange: geographic area of interest (one of six identified for field sampling)
- Latitude: latitudinal coordinate, in decimal degrees
- Longitude: longitudinal coordinate, in decimal degrees
- Aspect: direction of slope of plot
- Slope: average slope of plot, by ocular estimate, in percent
- % torch: proportion of tree canopy consumed by fire, averaged across all trees in plot area
- % scorch: proportion of tree canopy damaged, but not consumed by fire, averaged across all trees in plot area
- % mortality: proportion of trees in plot killed by fire
- average char height (m): maximum height along tree bole where bark was damaged by fire, averaged across all trees in plot area
- % consumption of heavy fuels: presumed percent reduction in 1000 hour Time Lag Fuel Moisture fuels on the ground, averaged across plot area
- torch_cbi: CBI value (ranging from 0 to 3) corresponding to % torch value assessed above
- scorch_cbi: CBI value (ranging from 0 to 3) corresponding to % scorch value assessed above
- mort_cbi: CBI value (ranging from 0 to 3) corresponding to % mortality value assessed above
- char_cbi: CBI value (ranging from 0 to 3) corresponding to char height value assessed above
- fuel_cbi: CBI value (ranging from 0 to 3) corresponding to % consumption of heavy fuels value assessed above
- cbi_field: total plot CBI value calculated by averaging five previous individual CBI values
- NA values signify attributes we couldn't assess in the field (eg. no scorch % if no live trees present pre-fire)
File: field_data_site.csv
Description: general site characteristics of each plot installed
Variables
- Site Name: identification code for field plot
- Date Visited: date of data collection
- Collected By: field sampling team
- Fire History: most recent fire which burned plot location, name and year
- Latitude: latitudinal coordinate, in decimal degrees
- Longitude: longitudinal coordinate, in decimal degrees
- x: longitudinal coordinate, in decimal degrees
- y: latitudinal coordinate, in decimal degrees
- Aspect: direction of slope of plot
- Slope: average slope of plot, by ocular estimate, in percent
- Vegetation Type: qualitative assessment of dominant vegetation in plot area
- Observations: notes on plot conditions and parameters, to add descriptive / qualitative data to dataset
File: field_data_treecount.csv
Description: summary of tree data for each plot
Variables
- Site Name: identification code for field plot
- Coordinates: latitude and longitude, in decimal degrees
- Aspect: direction of slope of plot
- Slope: average slope of plot, by ocular estimate, in percent
- Number Live Overstory: count of live trees in plot area, with DBH >= 12.7 cm
- Number Live Sapling: count of live trees in plot area, with DBH < 12.7 cm
- Total Live Tree: count of all live trees in plot area
- Number Standing Dead: count of standing dead trees in plot area, with DBH >= 12.7 cm
- Number Dead & Down: count of logs
To quantify tree mortality from contemporary fires, we sampled fire effects at 74 of the 406 fire history sites used in the study. We located the field sites across a gradient of contemporary burn severity and fire management strategies. For example, proportion of fires in the full suppression category ranged from 100% in the Pinaleño and Santa Catalina Mountains to 14% on the Kaibab Plateau. Similarly, the proportion of sites which burned with high probability of tree mortality (CBI > 1.61) ranged from 73% in the Santa Catalina Mountains to 0% on the Kaibab. Data collection focused on six key geographic areas where networks of fire history sites had been established prior to wildfires occurring over the past ten years (2011-2020): the Jemez, Rincon, Santa Catalina, Pinaleño, and Chiricahua Mountains, and the Kaibab Plateau. In the Jemez Mountains, sites are in Bandelier National Monument (including the Bandelier Wilderness), the Valles Caldera National Preserve, and the Santa Fe National Forest. The Rincon Mountain sites are mostly located in the Saguaro Wilderness (within Saguaro National Park), and about half of the sites on the Kaibab Plateau are in proposed wilderness in Grand Canyon National Park; the remaining sites sampled in Arizona are within the Coronado and Kaibab National Forests.
We relocated each fire history site and established a 10-m radius plot. If we found a tree or stump sampled in the original fire history data collection, the plot was centered at its location; if a sampled tree was not located (typically due to high-severity fire effects), we centered the plot at the coordinates provided by the original researcher. Where we found multiple sampled trees or stumps at least 20 m apart, we installed a plot at each, for a total of 91 plots at 74 distinct fire history sites. For all trees in the field plots, we recorded tree diameter at breast height (dbh), species, and status (live or dead). We measured diameter and assigned species for downed logs and recorded an overall count of trees both live and dead, standing and down. A qualitative description of site conditions and photographic documentation completed our site characterization.