2021 Fire Season Impacts to Giant Sequoias

Executive Summary

Dr. Kristen Shive*, Dr. Christy Brigham**, Tony Caprio**, and Paul Hardwick**
*The Nature Conservancy
**National Park Service, Sequoia and Kings Canyon National Parks

The 2021 fire season included two large wildfires (both started by the same lightning storm in early September) that burned into a large number of giant sequoia groves. This species has a limited distribution, covering just ~28,000 acres in ~70 groves on the western slopes of the Sierra Nevada mountains. Given the impacts of the 2020 Castle Fire to sequoia groves, where losses were estimated at 10-14% of the entire Sierra Nevada population of sequoia trees over 4 feet in diameter, there is significant concern by sequoia managers and the public regarding the impacts of these new fires. This report summarizes potential impacts to groves that were burned in the 2021 fire season, in the KNP Complex Fire and Windy Fire, including number of groves burned, amount of grove area burned at differing vegetation severity (RAVG-composite burn index), estimated number of large sequoias killed based on a preliminary analysis, the percentage of the entire population this loss represents, potential for regeneration failure, and potential for loss of seed source due to erosion (for KNP only). All of the data, data analysis, maps and modeling contained in this report are excerpted from the Burned Area Emergency Response Plan for the KNP Complex produced by an interagency team and submitted by the National Park Service. Some additional explanatory text and structure has been added to make this a standalone report. Because of this emphasis, there is more detail and analysis for KNP Complex sequoia groves than those burned in the Windy Fire.

Grove-level Estimates

• In total 27 sequoia groves are fully or partially within the fire perimeters of the KNP Complex Fire and the Windy Fire.

  • 16 groves burned in the KNP Complex Fire.

  • 11 groves burned in the Windy Fire.

  • For both fires combined, 6,109 acres of giant sequoia groves were burned. This estimate is based on updated grove boundaries provided by the NPS and the USFS Region 5 Remote Sensing Laboratory.

    • KNP: 4,374 acres

    • Windy: at least 1,735 acres (Note: this estimate does not include two groves on the Tule River Reservation for which there is not publicly available spatial data)

Sequoia Grove Vegetation Burn Severity Analysis Using Composite Burn Index (CBI)

Fire severity was assessed using the USFS Rapid Assessment of Vegetation Condition after Wildfire (RAVG) four class Composite Burn Index product (CBI4) (details on the severity maps can be found on the RAVG website https://fsapps.nwcg.gov/ravg/). This analysis contains four categories of vegetation change: undetected change, low, moderate, and high, where high severity has been associated with 95% tree mortality (Miller et al., 2009). The acreage burned by severity class for each fire is:

• KNP Complex (all ownerships):

  • Undetected change: 1,169 acres

  • Low: 1,849 acres

  • Moderate: 740 acres

  • High: 616 acres

• Windy Fire (all ownerships):

  • Undetected change: 228 acres

  • Low: 659 acres

  • Moderate: 437 acres

  • High: 411 acres

Large Sequoia Mortality Estimates

For the KNP Complex we estimate that between 1,330-2,380 sequoias over four feet in diameter have already been killed or will die within the next three to five years. This estimate is based on RAVG satellite analysis combined with mortality rates from surveys in other sequoia groves that burned in three previous fires (similar methodology to Stephenson and Brigham 2021).

For the Windy Fire, we estimate that between 931-1,257 sequoias over four feet in diameter have already been killed or will die within the next three to five years.

The combined impact of these two fires is estimated to be 2,261-3,637 sequoias over four feet in diameter that have already been killed or will die within the next three to five years. These losses make up an estimated 3-5% of the entire Sierra Nevada sequoia population over four feet in diameter. On top of the 10-14% of large sequoias lost in the 2020 Castle fire, these fire impacts represent a significant threat to large sequoia persistence.

Potential for Regeneration Failure

The least understood impacts of these wildfires are impacts to sequoia regeneration in high severity areas. Sequoias generally regenerate well after wildfire, though reports of inadequate regeneration in high severity areas are raising concerns. Regeneration failures could potentially occur if the cones and/or seeds were incinerated during crown fire, seeds did not survive the smoldering fire, or seeds washed away due to surface erosion. In these cases, regeneration would be dependent on proximity to live tree seed sources.

For the KNP complex we analyzed high severity areas within sequoia groves that were over 100 meters from an intact sequoia grove area with live sequoia trees (represented by areas of undetected change, low or moderate severity). In total 436 acres were identified that may be vulnerable to total sequoia loss if regeneration from seeds that survived the fire is inadequate. Of the 436 acres >100m from live sequoia forests, 335 acres are on NPS lands. The Burned Area Rehabilitation Plan recommends reforesting these 350 acres with giant sequoias if adequate regeneration is not present based on subsequent field surveys.

Erosion modeling suggests that the high severity areas identified for potential restoration are also at high risk of losing any seeds that did survive the fire due to surface erosion.

Other Key Points

In several places that burned during this event, previous prescribed fire work appears to have reduced fire severity (portions of Redwood Mountain Grove, Giant Forest). In other areas previous prescribed fire and mechanical thinning treatments, as well as preparation for upcoming burn units, allowed fire crews opportunities to safely fight fire more effectively (prescribed burn preparations at Lost Grove and Park Ridge Fire Lookout area were used during suppression operations on the KNP complex).

Although we are seeing some significant high and moderate severity areas in sequoia groves where we expect impacts to large giant sequoias to be detrimental, much of the grove area in the KNP Complex burned at low severity and we expect beneficial results for giant sequoias in these areas. These beneficial effects include fuel reduction, small canopy openings ideal for regeneration, and removal of litter and generation of ash – also ideal conditions for giant sequoia seedlings.

Other areas not classified as high severity may have also had beneficial effects, but the effects will likely be more mixed. In areas classified as “undetected change,” there will likely be a mix of completely unburned areas and areas that had a light surface fire that is similar to low severity fire effects. The fire effects in moderate severity will be the most variable, with some areas having beneficial effects and others being more severe.

Impacts of the 2021 Fire Season to Giant Sequoia Ecosystems

Extracted data and analysis from the NPS KNP Complex BAER report with additional text for context

Introduction

All of the data, data analysis, maps and modeling contained in this report are excerpted from the Burned Area Emergency Response Plan for the KNP Complex authored by the National Park Service. Some additional explanatory text and structure has been added to make this a standalone report. Because of this emphasis, there is more detail and analysis for KNP sequoia groves than those burned in the Windy Fire.

Both the two fires in the KNP Complex (Colony and Paradise) and the Windy Fire were ignited by lightning during a late evening thunderstorm on September 9, 2021. The KNP Complex burned 89,398 acres as of October 18, 2021 (78,675 acres in SEKI, 7,128 acres on Sequoia National Forest, 1,732 acres on BLM, 1,555 acres on private property, and 308 acres on state lands, with the fire remaining at 75% containment) and the Windy burned 97,528 acres by a containment date of Nov. 15, 2021. A full suppression strategy was utilized on both fires as soon as they were discovered.

Giant sequoias are the world’s most massive non-clonal organisms, storing immense amounts of carbon, globally ranking second only to coast redwood forests on a per acre basis (Sillett et al., 2015). Giant sequoia forests have a limited distribution, covering ~28,000 acres in ~70 groves across the western slope of the Sierra Nevada (Stephenson and Brigham, 2021). Because of their great size, age, and limited distribution, they have captured the public’s imagination and have been central to the origin of many state and national parks. They are specifically referenced in the enabling legislation for Sequoia National Park and are a focal resource within the park’s Foundation Statement. They are the premier destination within Sequoia and Kings Canyon National Parks (SEKI), attracting over 1.5 million visitors per year. SEKI has 38 delineated groves of giant sequoias, covering roughly 10,000 acres. Extensive surveys in the 1960s and 1970s, covering 90% of SEKI’s current grove areas, found 84,822 individual sequoia trees greater than 1 foot in diameter at breast height (DBH) and 25,181 greater than 3.5 feet DBH (following past convention we will round to the nearest foot and call the latter trees “greater than 4 feet”). All told, Sequoia and Kings Canyon National Parks manage approximately 37% of all existing sequoia grove area.

Like the mixed conifer forest matrix in which giant sequoia occur, historically fire was frequent and generally of low to moderate severity, with a historic fire return interval of ~15 years (Swetnam et al., 2009), which fluctuated with climate (Swetnam, 1993). This means that the oldest trees experienced many dozens of fires in their lifetimes. Small patches of high intensity also occurred and were important for sequoia regeneration, because they would thin the forest canopy and help open sequoia cones. Small openings created by localized high intensity fires varied between 0.1-0.4 hectares (0.25 acres to 1 acre) in size (Stephenson, 1994). With Euro-American settlement, lightning strikes were extinguished and indigenous burning was prohibited, resulting in high levels of surface fuels and increased tree densities in many groves (Kilgore and Taylor, 1979; Parsons and DeBenedetti, 1979; Stephens et al., 2015). As early as the 1960s, sequoia managers realized the importance of fire and began reintroducing fire in 1969 into several groves with prescribed fire and managed wildfire (Kilgore, 1970). However, resource limitations, as well as policy, boundary limitations, and planning impediments, have limited their implementation Botti and Nichols, 2021). Of the ~10,000 acres of giant sequoia groves in SEKI, 4,610 acres have received one or more prescribed or managed wildfire in the last 20 years, mostly in Giant Forest.

High Severity Fire and Giant Sequoias

Giant sequoia are highly fire-adapted, but they are not adapted to extensive high severity fire. They cannot sprout after crown loss, and recent work suggests that large sequoia are unlikely to survive with >~85% crown damage for trees with one or more existing fire scars and >~90-95% crown damage for trees without existing fire scars (Shive et al., in review).

Since 2015 there has been a dramatic increase in area burned and amount of high severity fire in sequoia groves, resulting in significant mortality of large, legacy sequoias (Shive et al. in review; Stephenson and Brigham, 2021). This increase is consistent with increases in high severity fire throughout the western US, which has been linked with warmer, drier fire seasons (Parks and Abatzoglou, 2020) and increasing fuel aridity linked with climate change (Williams et al., 2019). The warming climate is also directly impacting forests via “hotter droughts” (Williams et al. 2015). The hotter drought of 2012-2016 resulted in widespread conifer die-off (Young et al., 2017), which could be increasing fire severity in this region (Wayman and Safford, 2021). The interaction of these climate-driven trends with elevated fuel loads, a result of fire exclusion, is increasingly putting sequoia groves at risk of severe fire (Kilgore and Taylor, 1979; Parsons and DeBenedetti, 1979; Stephens et al., 2015).

Surveys in old growth groves that burned in three wildfires from 2015 to 2017 documented an average mortality rate of ~84% in high severity areas, which includes delayed mortality three to five years postfire (Shive et al., in review). In 2020, the Castle Fire burned 9,531 acres of giant sequoia, 2,810 of it in high severity. In total, it is estimated that 10-14% of all existing large sequoias (>4’) across the range were killed as a result of the Castle Fire (Stephenson and Brigham, 2021). Significant amounts of high severity fire effects were observed within SEKI boundaries, most of that concentrated in Board Camp, Upper Dillonwood, and Homer’s Nose groves.

In addition, there are anecdotal reports of regeneration failures in many high severity areas. Regeneration failure could be the result of lack of seed survival, lack of seed dispersing into the site, or failure of seedlings to establish because of environmental stress. Heat from fire typically dries out cones for a mass seed release that is timed for ideal postfire environments – bare mineral soil, an ash layer and canopy openings (Hartesveldt et al., 1975). Because of the relationship between seed release and fire, scientists and managers had hoped that though there were major losses of the ancient trees in high severity areas, that regeneration would be prolific. Preliminary data from Black Mountain Grove (2017 Pier Fire) suggests that this is the case in many high severity areas. However, in the preliminary data from Nelder Grove, where there was significant crown torching, the highest regeneration was documented in moderate severity areas with minimal regeneration in the highest severity area (Amarina Wuenschel and Andrew Latimer, unpublished data). Preliminary field surveys by NPS staff in Board Camp Grove (2020 Castle Fire) suggest similar patterns (C. Brigham personal communication). Regeneration failures have also been noted in high severity areas in locations with high heat loads, such as south or west-facing slopes (Tony Caprio, personal communication).

The KNP Complex burned a total of 4,374 acres of sequoia groves, of which 616 acres burned at high severity. Simultaneous with the KNP Complex, the Windy Fire also burned 1,723 acres of giant sequoia groves in the Giant Sequoia National Monument, of which 411 acres was high severity.

Due to concerns about the changing impacts of wildfire on forests and particularly on giant sequoia forests, this report was prepared utilizing data, text, and analyses from the NPS KNP BAER report to provide initial estimates of large sequoia mortality from these two fire events. These estimates will be used to inform sequoia managers and the general public regarding impacts of these fires to giant sequoias as well as to target sequoia mortality field surveys and plan further research and management to address fire impacts to this important species.

METHODS

Sequoia Grove Vegetation Burn Severity Analysis Using Composite Burn Index (CBI)

Fire severity was assessed using maps from the USFS Rapid Assessment of Vegetation Condition after Wildfire (RAVG) program which are created by the USFS Geospatial Technology and Applications Center of the US Forest Service. RAVG uses remote sensing to estimate burn severity on a given fire.

RAVG maps are based on pre- and post-fire comparisons of Landsat imagery (30x30 meter pixel resolution). For our analysis we used the burn severity classification of the Composite Burn Index (CBI), a standardized fire severity rating based on a composite of effects to the understory vegetation (grass, shrub layers), midstory trees and overstory trees (Background, Products & Applications | Rapid Assessment of Vegetation Condition after Wildfire (RAVG) (nwcg.gov). We selected the CBI product over others (such as basal area loss or canopy cover loss) as it is consistent with Stephenson and Brigham (2021). The CBI is classified into severity classes based on existing regression models, and these severity classes have in turn been linked with percent vegetation change (Miller and Thode, 2007; Miller et al., 2009). This results in four fire severity classes: undetected change (0% change); low (1-25% change); moderate (26-75% change); high (76-100% change). Though the high severity class has been broadly assessed as 76-100% change in live vegetation, models comparing field plots have linked this class with 95% tree mortality (Miller et al., 2009).

For the KNP Fire, the pre-fire image was from September 8th, 2021 and the post-fire image was from October 13th, 2021. For the Windy Fire we used a preliminary map, where the pre-fire image was from October 7th 2020 and the post-fire image was from October 10th, 2021.

Large Sequoia Mortality

To obtain estimates of individual large sequoia loss (defined as sequoias >4 feet in diameter), two methods were employed. First, for areas for which individual stem maps of sequoia trees giving their size and location were available (burned groves of Sequoia and Kings Canyon National Parks, Whitaker’s Forest, and Save the Redwoods League properties), fire severity maps were overlaid with stem maps to identify how many sequoias of a certain size were located within the areas burned at differing severities (all estimates are based on >4 feet in diameter trees; for NPS groves we also analyzed a subset of this class, very large sequoias, those with a diameter greater than 10 feet). To estimate mortality of these trees, we used datasets from three wildfires (2015-2017) of post-fire giant sequoia mortality (Shive et al., in review). These field surveys include delayed mortality 3-5 years post-fire and show a wide range of mortality for low, moderate, and high severity fire areas. To reflect this variability, we calculated a range of possible losses, using the highest and lowest loss rates for each severity class (see Table 1). These fractional losses by severity classes were then multiplied by the number of sequoia trees of a given size in each fire severity class to provide a high and low estimated loss.

Table 1. Mortality rates observed by Shive et al. (in review) for the Rough and Pier Fires. The rates for the Railroad Fire used in this assessment are from Stephenson and Brigham (2021).

Survey area	Low	Moderate	High
2015 Rough Fire (Evans, Kennedy and Lockwood Groves)	0.0%	14.0%	75.0%
2017 Railroad Fire (Nelder Grove)	5.9%	22.2%	100.0%
2017 Pier Fire (Black Mountain Grove)	11.9%	24.5%	74.6%

* The surveys for the Pier Fire and Rough Fire were focused on moderate to high severity. In these fires, observations in low severity were immediately adjacent to moderate and high severity, which could somewhat elevate the mortality rate. However, the Rough Fire has a 0% mortality rate in low severity so we feel it is sufficiently conservative (Shive et al., in review). For the Railroad Fire, we used the mortality rates presented in Stephenson and Brigham (2021), which are based on additional surveys outside of the Shive et al. (in review) study (unpublished data, Amarina Wuenschel, USFS), because they had a larger sample size in low severity and were overall more conservative.

For areas where no stem maps exist (Giant Sequoia National Monument, Tule River, and private lands) an average per acre value for sequoias over 4 feet in diameter derived from the NPS Sequoia Tree Inventory (2.61/acre) was used to estimate the number of large sequoias present in each area by fire severity class (Stephenson and Brigham 2021). This estimated number of large sequoias was then multiplied by the same mortality percentages as above to generate a high and low estimate of large sequoia loss. One note on the actual measured diameters, here we are using large sequoia numbers from the Sequoia Tree Inventory (STI) where diameters were recorded in one-foot classes and our minimum size (4’) was captured in the 3.5-4.5’ diameter class. Following past convention we round to the nearest foot and call these sequoias “four feet or larger” (Stohlgren 1991).

We did not separate out immediate versus delayed mortality here, because we only have data from one fire to support that distinction. We do note however, that for the 2017 Pier Fire, of the total number of dead trees by year three (72), ~71% were considered initial mortality, while the remaining were dead by year three post-fire. Finally, we also present mortality estimates in the context of the estimated total population of large sequoias across the range, which was estimated by Stephenson and Brigham (2021) to be ~75,580.

The caveats and uncertainties of this methodology discussed in Stephenson and Brigham (2021) apply to this analysis as well.

Potential Regeneration Failure

The least understood impacts of these wildfires are impacts to sequoia regeneration in high severity areas. Sequoias generally regenerate well after wildfire, though reports of inadequate regeneration in high severity areas are raising concerns. Regeneration failures could potentially occur if the cones and/or seeds were incinerated during crown fire, did not survive the smoldering fire, or the seeds washed away due to surface erosion. In these cases, sequoia regeneration would be dependent on proximity to live trees as a future seed source.

If seed loss or seedling failure occurs in a large, contiguous area, then sequoia recolonization must occur via seeds generated by nearby living giant sequoias. To understand how much high severity giant sequoia area might be outside of the dispersal range for live trees, we conducted a spatial analysis to identify the number of acres that are >100m from live tree seed sources. This is a conservative measure of seed dispersal distance, because although no comprehensive studies of seed dispersal in giant sequoia exist, Harvey, Shellhamer, and Stecker (1980) found a young sequoia established 89 meters beyond the nearest large sequoia. Additionally, since some heterogeneity and patchiness within sequoia groves is to be expected and restoration efforts, if needed, must be targeted, we only analyzed high severity patches over 10 acres in size.

Surface Erosion Modeling FS WEPP-PEP

To evaluate soil erosion and sediment potential in giant sequoia groves, modeling usingthe WEPP (Watershed Erosion Prediction Project) Cloud, post-fire erosion prediction (PeP) model was performed on select groves https://wepp.cloud/weppcloud/runs/hygroscopicwaste/baer/ ). The model predicts hillslope erosion and deposition, channel erosion, and totalsediment discharge for a catchment based on inputs and reports erosion rates by hillslopeand/or channel (Elliot 2006). The inputs that WEPP-PeP utilizes are:

• USGS DEM and Land Use Layers
• NRCS SSURGO and STATSGO soil databases
• PRISM climate information
• SBS map uploaded for the fire perimeter area

This model was used exclusively to illustrate where within the giant sequoia groves will be at most risk of elevated post-fire erosion.

RESULTS

Patterns of severity

KNP Complex

A total of 4,374 acres of grove area burned in the KNP Complex, which included portions of 16 distinct groves. In the limited areas we were able to ground truth or observe via helicopter, the areas designated as “undetected change” did have an underburn; more extensive surveys are needed to confirm the extent of this pattern. Low severity fire was likely largely restorative and it represents the largest proportion by severity class of total grove area burned (42%). While undetected change within the fire perimeter and low severity represent 70% of the burned grove area, the absolute acreage of high and moderate severity, where most mortality of large sequoias is expected to occur, is significant (Tables 2a and 2b, Figure 1).

Table 2a. Acres burned by sequoia grove and severity.
Total grove acres burned in KNP: 4373.6

Grove	Outside Fire Perimeter	Severity				Total burned area
		Unchanged	Low	Moderate	High
Atwell	601.4	52.9	197.1	51.7	18.9	320.6
Big Springs	0	0	2.2	0.1	0	2.3
Castle Creek	0	44.2	122.6	48.7	2.4	217.9
Douglass	0	0.1	0.9	0	0	1
East Fork	739.4	4.2	1.4	0	0	5.6
Giant Forest	1609	284.8	183.3	24.0	3.7	495.8
Lost	4.8	17.6	12.3	0.3	0	30.2
Muir	0	69.9	139.7	28.5	6.8	244.9
New Oriole Lake	0	0	7.8	3.6	3.1	14.5
Oriole Lake	0	49.4	90.9	6.1	0.4	146.8
Pine Ridge	0	0.6	39.4	3.1	0	43.1
Redwood Creek	2.3	5.8	32.7	4.6	1.6	44.7
Redwood Mountain
NPS	0	549.9	656.7	411.3	457.0	2074.9
USFS	0	7.5	113.6	78.2	95.4	294.7
UCB	0	74.3	187.4	30.2	8.2	300.1
Skagway	0	7.4	47.6	4.5	1.3	60.8
Squirrel Creek	0	0	6	0.9	0	6.9
Suwanee	0	0.1	7	43.8	17.5	68.4
Total acreage		1168.7	1848.5	739.6	616.3	4373.6
Percent of total		27%	42%	17%	14%

Table 2b. Percent of area burned by severity (including percent outside fire perimeter) by sequoia grove. Some rows don’t sum exactly to 100% because we rounded to whole numbers.

Grove	Outside Fire Perimeter	Severity
		Unchanged	Low	Moderate	High
Atwell	65%	6%	21%	6%	2%
Big Springs	0%	0%	96%	4%	0%
Castle Creek	0%	20%	56%	22%	1%
Douglass	0%	10%	90%	0%	0%
East Fork	99%	1%	0%	0%	0%
Giant Forest	76%	14%	9%	1%	0%
Lost	14%	50%	35%	1%	0%
Muir	0%	29%	57%	12%	3%
New Oriole Lake	0%	0%	52%	24%	21%
Oriole Lake	0%	34%	62%	4%	0%
Pine Ridge	0%	1%	91%	7%	0%
Redwood Creek	5%	12%	70%	10%	3%
Redwood Mountain
NPS	0%	26%	32%	20%	22%
USFS	0%	3%	39%	27%	32%
UCB	0%	25%	62%	10%	3%
Skagway	0%	12%	78%	7%	2%
Squirrel Creek	0%	0%	86%	13%	0%
Suwanee	0%	0%	10%	64%	26%

Figure 1. Map of groves showing fire severity for KNP Complex Fire.

Redwood Mountain Grove had the most significant acreage impacted by high severity fire (Figure 1, 2a, 2b), followed by Atwell and Suwanee. Where Atwell burned severely, the forest was mostly second growth. Reconnaissance via helicopter over Suwanee indicated that many of the areas classified as moderate severity were underpredicted in the severity map, potentially resulting in more mortality than estimated here.

Figure 2. Redwood Mountain Grove. Views of (a - above) large high severity patch at the southern end of the grove from 10/12/2021 and (b - below) a close-up view of dead sequoias from 10/27/2021.

Windy Fire

Total sequoia grove area burned where public data was available was 1,735 acres. This does not include any groves fully on tribal lands but does groves that are on both USFS and tribal lands as well as private property within the burn area. These values are different from those reported by USFS because we used newly updated grove boundaries produced by USFS Region 5 remote sensing lab and NPS-SEKI. These boundaries are still in draft form, but field observations suggest they are much more precise than prior boundaries (Stephenson and Brigham 2021). Burn severity using the four class product was as follows (Fig. 3):

• Undetected: 228 acres
• Low: 659 acres
• Moderate: 437 acres
• High: 411 acres

Large sequoia mortality

KNP Complex

For groves on NPS land we found significant losses in both size categories, 4 to 10 feet in diameter and greater than 10 feet in diameter. The majority of these losses were in Redwood Mountain Grove. In total, our estimates suggest that roughly 1,105-2,020 large sequoias were either killed directly by the fire or will experience delayed mortality within 3-5 years postfire (Table 3) on NPS land. This represents ~1-3% of the range-wide population of large sequoias (Stephenson and Brigham, 2021).

Table 3. Estimated mortality in the KNP Complex on NPS land.

Size Class	Low	Moderate	High	Total
4-10 feet
Minimum estimated mortality	0	121	464	585
Maximum estimated mortality	262	201	619	1,082
>10 feet
Minimum mortality	0	99	421	520
Maximum estimated mortality	214	164	560	938
All >4’
Minimum estimated mortality	0	220	885	1,105
Maximum estimated mortality	476	365	1,179	2,020

We also estimated mortality in 594 acres of Redwood Mountain Grove that are outside of SEKI’s boundaries. For the 300 acres that burned on Whittaker’s Research Forest (which is managed by UC Berkeley) we used an existing stem map that included 233 large sequoias (>4 feet), similar to the approach used above. This portion of the grove burned at primarily lower severity, with estimated mortality of 10-26 large sequoias. Staff at UC Berkeley plan to do a field assessment of mortality in summer 2022.

For the 294 acres in the Giant Sequoia National Monument, we did not have access to stem maps so we used the mean density of large sequoias per acre from the STI (~2.61 per acre) to estimate the number of large sequoias on the landscape. We then used the mortality rates from Table 2, estimating 215-334 large sequoias may have been killed on these acreages. Across all grove ownership on the KNP Complex, we estimate that 1,330-2,380 sequoias over four feet in diameter have already been killed or will die within the next three to five years.

Windy Fire

We also estimated the impacts of the 2021 Windy Fire, which burned through 1,735 acres in 12 groves that are predominantly in Giant Sequoia National Monument, with smaller acreages managed by Save the Redwoods League (SRL), the Tule River tribe and other private landowners.

SRL had stem-mapped their portion of the Red Hill Grove, yielding 118 total large sequoias on their property. Applying mortality estimates to those by severity class, 16-29 were either immediately killed by the fire or are expected to die within three to five years (Table 5). For all other lands we used the average density of large sequoias from SEKI (2.61 sequoias >4 ft DBH per acre) (similar to above for the KNP Complex on GSNM lands) to estimate a loss of 915-1228 sequoias over 4 feet in diameter (Table 4). We note that for the Tule River reservation, our estimate includes groves in the publicly available boundary dataset with acreage on both Tule and USFS lands (Black Mountain, Peyrone, South Peyrone and Red Hill), but our estimate does not include any other grove acreage that is wholly on the Tule River reservation. Combining SRL lands with all other lands where we have grove spatial data, an estimated 931-1,257 sequoia trees over four feet in diameter may be lost as a result of the Windy Fire.

Table 4. Mortality for non-SRL lands (where there are no stem maps)

Windy (non-SRL)	Severity class			Total
	Low	Moderate	High
min	0	147	768	915
max	193	253	782	1228

Table 5. Mortality for Windy Fire – Stem mapped SRL Lands

Windy (SRL only)	Severity class			Total
	Low	Moderate	High
min	0	7	9	16
max	7	11	11	29

Combined Mortality Estimates

Combining both NPS and non-NPS lands burned in the KNP Complex with the Windy Fire, ~2,261-3,637 large sequoias are estimated to have been either killed outright or will die within a few years. This suggests that the 2021 fire season resulted in a loss of ~3-5% of the large sequoia population. When considered with mortality from the 2015-2017 fire events and the estimated loss of 10-14% of large sequoias in the 2020 Castle Fire, this fire season contributes to an alarming trend.

Potential Regeneration Impacts in the KNP Fire
Regeneration failure in larger areas of high severity fire effects in giant sequoias groves is a new and little understood phenomenon. As pointed out previously in this document, giant sequoias typically reproduce well after fire and depend on many aspects of post-fire conditions to stimulate germination and increase seedling survival (Harvey et al. 1980). However, where seeds do not survive the fire due to severe fire effects, areas that are large and distant from sequoia seed source are at possible risk of regeneration failure (either due to seed loss or seedling death). Of the 207 distinct patches of high severity fire effects in giant sequoia groves in the KNP fire, only five patches are >10 acres. Where the patches were smaller, they may have contributed to heterogeneity. Two areas of Redwood Mountain Grove had 436 acres >100m from a likely surviving giant sequoia seed source, 353 acres of that are in SEKI (Figure 4). Within these high severity areas of Redwood Mountain Grove, some very small patches of moderate severity located within these high severity patches were excluded, because drone imagery suggests that these areas may be fully or near-fully scorched (Paul Hardwick, personal communication, Figure 5). Active reforestation may be needed in this area for giant sequoia, with the extent of this need determined by field surveys in 2022. The NPS BAER plan contains a funding request for this restoration action should giant sequoia regeneration failure be documented in these areas during subsequent field surveys.

Figure 4. Southern end of Redwood Mountain Grove with significant patches of high severity.

Figure 5. Ortho-rectified drone imagery of small moderate severity patches within the large high severity patch of Redwood Mountain Grove. Because surviving trees in these patches could be limited, the acreage calculation for distance to live tree excluded these for the regeneration analysis (Paul Hardwick, personal communication). Orange squares are classified as moderate severity in RAVG.

Soil Erosion Modeling (FS WEPP-PEP)
Soil erosion modeling indicated that several groves may experience erosion significant enough to remove seeds located on the surface of the soil. Large areas of Redwood Mountain Grove (Figure 6) and Skagway Grove (Figure 7) appear subject to surface erosion. Where these areas overlap with large areas of high burn severity, grove regeneration may be compromised.

Figure 6. Soil erosion potential in Redwood Mountain Grove.

Figure 7. Soil erosion potential for groves near Lodgepole area.

Figure 8. Soil Erosion Potential for Groves Near Mineral King.

Giant Sequoia Groves – Burn Narrative Details

We also provide a narrative description of fire behavior and timing for each grove as this information may, in part, help understand patterns of burn severity.

A total of 16 groves had some portion of the grove burned by the KNP Complex. In some grovesthis was the whole grove and, in some cases, just a small portion. Groves were located onNPS (SEKI), USFS (GSNM), and UC Berkeley (Whitaker’s Forest) lands. Additionally, there werefive groves where some kind of fire suppression activity or pre-fire prep work around giantsequoia trees took place. The latter groves included Case Mountain (BLM’s only sequoia grove),Big Stump (NPS and USFS), Grant Grove (NPS), Bearskin (USFS), and Redwood Meadow(NPS). Plans for the aerial application of fire-retardant GEL were developedfor six groves, Suwanee, Muir, Castle Creek, Redwood Mountain, Big Stump, and RedwoodMeadow, but were only implemented in a portion of Muir Grove.

Oriole Lake – 9/16 to 9/26 – Fire flanked and backed into the grove from the northwest off of Paradise Ridge burning into the lower portions of the grove first and slowly moving into highareas, further east up Squirrel Creek, over the next 10 days.

Suwanee – 9/17 – All of Suwanee burned during an extreme headfire run north up the MarbleFork drainage that burned about 6,500 ac. The bulk of the grove burned with high or moderateseverity with a couple of small low severity patches.

Giant Forest – 9/17 to 10/18 – The main portion of the KNP Fire burned into the westernareas of Giant Forest during the run on 9/17, primarily burning into lower Deer Creek. On the18th the fire pushed further east into the grove burning across Generals Highway and the Crescent Meadow Road on the west side of the grove and into Sherman Creek area on thenorthwest side of the grove with one spot fire located about 130 meters northwest of the Sherman Tree on the east side of Generals Highway. On neither the 17th or 18th did the fire expand to any degree within the recently completed (2019) Rx unit west of Generals Highway. Fire continued to expand along this flank into the grove on 9/19. Burnout operations were implemented by Arrowhead and Alpine Hotshots, and the Whiskeytown module on 9/19 on the west and southwest flanks of the fire to prevent further spread of high severity fire into Giant Forest. On 9/21 burnout operation implemented in upper Sherman Creek and down around Pinewood. On 9/22 burnout operations were carried out in the Sunset Rock area. Fire continued to flank east around the south side of the grove toward Moro Rock (where a historic restroom burned on October 3rd) with burnout operations continuing ahead of the fire within the grove and south of the Crescent Meadow Road and High Sierra Trail into late October as a precaution in case the fire downslope of Giant Forest made an uphill run. Drought mortality made containing these burnouts difficult. No impacts to trees of special interest have been reported for Giant Forest other than the charring of the bases of two of the Four Guardsmen.

Squirrel Creek – 9/18 – Fire backed into these few trees along Squirrel Creek near the OrioleLake Road crossing.

Douglass – 9/22 to 9/23 – This very small grove on the north side of Paradise Ridge in theParadise Creek drainage appeared to have burned with a headfire.

Redwood Creek – 9/23 to 10/2 – Fire approached and backed into the grove from across Conifer Ridge until 10/1 when it ran to the east out of Redwood Creek drainage up toward Atwell Grove. Portions of the grove below Mineral King Road did not burn.

New Oriole Lake – 9/24 to 9/25 – Most of the grove burned on 9/24 after the fire made a pushout of the Oriole Lake area, east of where fire suppression activities were occurring. The grove burned under an inversion.

Skagway – 9/24 to 9/29– backing fire off Pine Ridge. Grove map does not show the grove extending up to the top of Pine Ridge but ground observation (Tyler Schmitt personal communication) and photos from helicopter recon indicate this is the case.

Muir – 9/25 to 10/5 (east flank) – A firing operation was carried out on 9/25 over the upperridgetop with the fire then backing off the ridge to east, west and north. Significant portions of the grove burned on 9/25, 9/30 (lower west area), 10/1 (west center and north area). Areas burned on 9/30 and 10/1 appeared to have been burned by a fire burning upslope from thelower canyon (possibly out of Pine Ridge Grove). A plan for aerial application of fire retardantGEL was prepared for this grove that focused on application to sequoia trees alongthe lower and upper border of the grove, depending on how fire spread into the area, topotentially reduce severity of a fire burning upslope into the grove. It was reported that GELwas applied to the grove. Documents provided by the Incident Management Team provide a point location within the grove and volume of gel applied. Post-treatment monitoring is being developed. Report from fire personnel on the ground was that when the fire hit the zone of gel application it dropped to the ground.

Pine Ridge – 9/26 to 9/30 – Appeared to have been largely burned by backingfire coming off Pine Ridge. Fire severity was predominantly low with some moderate.

Castle Creek – 9/26 to 10/2 – The fire approached the grove from the west flanking into thegrove over several days from 9/26 to 9/29. On 9/30 the fire made a strong push to theeast below the grove and then upslope into the lower portions of the grove, burning lowerelevation pockets of trees. On 10/1 this push continued with the fire burning upslope throughthe grove to the south and east. It was reported that the sun got on this grove resulting ina headfire while burning upslope. A column of unknown size was reported. This grovepreviously burned in a backing fire during the 1996 Castle Fire.

Atwell – 9/30 to 10/18 – The fire first flanked into the grove from the west on 9/30 andfollowed on 10/1 by a major run into the west side of the grove from the Redwood Creek area.Fire continued to flank through the grove to the east and crossed Mineral King Road to thesouth. One 10/5 the fire made a significant uphill run from south of the road up into the centerof the grove resulting in some high severity effects. Much of the area that burned in this part ofthe grove was second growth forest that had been logged prior to the area being purchased and incorporated into the parks. The fire continued to flank east through the grove, although spread was minimal into higher elevations by this date. The fire made one last push upslope below the Atwell Campground on about 10/18 along the west boundary of the 2016 Deadwood Rx. Condition of the four trees of special interest in the grove is unknown.

Big Springs – 10/2 to 10/3 -- The three-small pockets of trees that make up this grove burned on 10/2 and 10/3 under an inversion.

Redwood Mountain – 10/4 to 10/9 -- The initial burning in this grove was blacklining operationsthat were initiated at night along Generals Highway on 10/1 (southwest side of highway in areaof Kings Canyon Overlook). These operations continued 10/2 with crews working southeast andnorthwest along the highway from the previous area burned. By the morning of 10/3 the areafrom Big Baldy Trailhead to Redwood Canyon Overlook had been completed. On the evening of10/3 operations continued south along the Big Baldy Trail and west from the overlook to Redwood Ridge. During all three nights the Arrowhead UAS was used for recon and to conduct interior firing by dropping balls to increase blackline depth. On the morning of 10/4 they werefiring out on the east flank of Redwood Ridge south of Redwood Saddle. Late in the evening of10/3 fire that was burning to the west of Redwood Mountain in the Pierce Meadow area appears to have begun a run up the west side of Redwood Mountain Ridge (based on distant IR drone images and burn scar patterns). On the morning of 10/4 this fire was running across the ridge into the main Redwood Canyon and appears to have merged with another run that came up-canyon from the area that was burning just north of the Big Springs Grove. These runs merged and burned the southern end of Redwood Mountain Ridge and lower portions ofthe grove in Redwood Canyon at high severity and ran to the east, just to the south of the EastFork of Redwood Creek onto the Big Baldy Ridge. The fire also ran to the north through the Redwood Canyon until reaching the blackline burn but the severity was moderated by previous Rx burns (2009 Hart Rx, 2011 Redwood Rx, 2012 Whitaker Rx, and 2016 Goliath Rx). Area burned on 10/4 was 11,678 ac. The Redwood Saddle Cabin and Barton’s Cabin Log burned on this day (condition of the Tunnel Log, Roosevelt and Hart Trees are unknown). Much of the 2012 Rx did not immediately burn but the fire flanked west from the Redwood Saddle and burned into the Whitaker’s Forest area and late in the day made another run north across the Generals Highway and into GSNM lands near Bacon Meadow, east of Park Ridge and northwest of Quail Flat. The fire then continued to back to the west into the Eshom area on USFS lands, burning nearly all the remaining portion of the grove.

Lost – 10/4 to 10/5 – Backing fire was put into grove on 10/4 and was immediately followed by a headfire run just to the northwest of the grove that just missed the grove.

East Fork – 10/8 to 10/9, and 10/18 – Only a small section of the East Fork Grove on the north side of the East Fork burned over three days (prior to the recent remapping of sequoia grove boundaries sequoia trees on the south side of the East Fork were considered within the EastFork Grove and trees on the north side of the river as in the Atwell Grove).

DISCUSSION

Overall, the KNP Complex and Windy Fires burned all or portions of twenty-eight sequoia groves, burning a total of 6,109 grove acres out of an estimated ~28,000 grove acres rangewide. Although much of this acreage burned at undetected change to low severity (3,905 acres) and is expected to have beneficial effects on grove ecosystem functioning, a total of 2,204 acres burned at moderate to high severity. In previous recent wildfires that burned at moderate to high severity we have seen significant mortality of large giant sequoias (Shive et al., in review). For these two 2021 fires our preliminary estimates (based on mortality rates from these previous field surveys combined with severity mapping for these two fires) suggest a potential loss of giant sequoias over four feet in diameter between 2,261 and 3,637 large giant sequoias. These estimates need to be updated by field surveys in coming years to document the full impact of these fires. Current maps of fire severity and estimated losses from this report can be used to stratify field survey efforts across severity and anticipated mortality.

The findings in this report indicate that wildfires that burn under conditions that result in high to moderate severity fire effects are a significant threat to the persistence of large sequoias. Fires burning with large areas of high severity is a dramatic change from historic fire patterns. Data from previous prescribed burns, wildfires, and tree-rings indicate that prior to the impacts of climate change and fire exclusion, large numbers of large giant sequoias were not killed during fire events (Stephenson 1996).

Our analysis of the KNP Complex, as well as the Windy Fire Burned Area Emergency Response report, indicate that prescribed fire and thinning treatments can reduce fire severity and provide fire fighters with opportunities to safely control and manage wildfires in some locations under some conditions during wildfire events. These treatments may not be effective under all wildfire conditions but did appear to positively affect fire behavior and allow fire suppression in Giant Forest and other groves impacted by these fires.

The mortality values within this report are estimates of potential mortality. The mortality rates used for moderate severity fires do not reflect the highest mortality rate measured in post-fire sequoia groves (45% in Save the Redwoods League data for Nelder Grove) because we wanted to be conservative in our estimates. Moderate severity areas in particular should be tracked over time to assess how these rates vary by location over time.

The KNP Complex BAER report suggests a potential management action of replanting giant sequoias in high severity areas greater than 100 meters distant from intact sequoia grove areas. While planting of giant sequoias as part of reforestation, restoration, and plantation forestry has been done successfully in California (see project descriptions from Sierra Pacific Industries at https://spi-ind.com/ and Fahey et al. 2012 and references therein), evaluating whether post-fire seedling densities are sufficient to restore giant sequoias to burned areas should be done prior to moving ahead with active reforestation. In addition, areas proposed for reforestation should be evaluated for their potential to persist under a changing climate. The National Park Service memo regarding climate change adaptation and the framework focusing on Resist, Accept, Direct (NPS 2012) and the USFS General Technical Report 270 (Meyer et al 2021) are useful frameworks for evaluating restoration areas within a climate change context.

Finally, these ecosystems will continue to see wildfire. Grove areas that were in the fire perimeters but that had very little or no wildfire (“undetected change”) may still be at risk of severe fire in the immediate future. In areas where surface fuels were removed and tree densities were reduced, reburning is unlikely for ~10 years, giving ample opportunity to plan for the next wildfire or prescribed burn. Although reburning at low severity would be desirable, high severity reburns are of concern. Many past high severity burn areas in mixed conifer forests, which are a similar fuel type, have reburned severely due to the high fuel loads created by dense fire-killed trees that eventually fall to the surface, and vigorously regenerating shrubs (Coppoletta et al., 2016; van Wagtendonk, 2012). The potential for reburning at high severity in sequoia groves should be a priority for field investigation and where such an outcome seems likely, these areas should be targeted for fuel reduction work.

RECOMMENDATIONS FOR FUTURE WORK

1. Accurately report to the public what happened to sequoia groves during this fire event by reporting current mortality estimates and field verifying these estimates during 2022-2025 (see #3).

2. Consider reforesting 350 acres of giant sequoia groves on NPS lands that are vulnerable to regeneration failure (restoration dependent on results of field evaluations of natural regeneration in these areas).

3. Monitoring & Research

a. Track individual sequoias for at least three years for delayed mortality, including documentation of cedar bark beetle activity.

b. Survey high severity areas to assess regeneration patterns in Redwood Mountain Grove in summer of 2022 to inform potential sequoia planting.

c. Survey high severity areas to assess regeneration patterns in Redwood Mountain Grove to better identify the drivers of regeneration patterns specific to high severity areas that have not been investigated to date, including differences in scorched vs torched stands and site conditions.

d. Investigate mortality and survivorship from this event so that these response variables can be correlated with other predictor variables (slope, aspect, management history, fuel loading). Knowledge gained will be used to prioritize other sequoia groves for fuel reduction treatments before they are impacted by high severity fire.

e. Fully assess (or design stratified random subsample) fire effects in sequoia groves in spring and summer 2022.

f. Model connections between pre-fire fuel-loading, management history, weather, slope, aspect and other relevant factors and fire effects observed in burned sequoia groves during the recent fires that burned sequoia groves including Castle, KNP, and Windy. Use this model to maintain inform risk management actions to address high severity fire in sequoia groves.

g. Assess fuel loading (both standing dead and large woody debris) in current groves, including those that burned at low and moderate severity, to inform future treatment plans.

h. Assess potential for grove areas to reburn at high severity in the next 10 years to inform future treatment plans.

i. Develop and implement a plan to examine effects of retardant gel application on trees, other vegetation, soil, and water.

j. Generate mechanistic hypotheses to explain fire, drought and beetle interactions and continue to test pre-burn treatments to eliminate physiological damage that results in vulnerability to beetle attack.

k. Develop a strategy with partners for characterizing and communicating “good” fire versus “bad” fire, especially with respect to giant sequoias.

REFERENCES

Botti, S., and Nichols, T. (2021). National Park Service fire restoration, policies versus results: What went wrong. Parks Stewardship Forum 37.
Coppoletta, M., Merriam, K.E., and Collins, B.M. (2016). Post-fire vegetation and fuel development influences fire severity patterns in reburns. Ecol. Appl. 26, 686–699.
Fahey, C., R.A. York, and T. Pawlowska. 2012. Arbuscular mycorrhizal colonization of giant sequoia (Sequoiadendron gigateum) in response to restoration practices. Mycologia, 104(5) pp. 988-997.
Harvey, H.T., Shellhammer, H.S. and Stecker, R.E. (1980). Giant Sequoia Ecology. (Washington, DC: U.S. Department of the Interior, National Park Service).
Hartesveldt, R.J., Harvey, H.T., Shellhammer, H.S., and Stecker, R.E. (1975). The giant sequoia of the Sierra Nevada (Washington, DC: U.S. Department of the Interior, National Park Service).
Kilgore, B.M. (1970). Restoring Fire to the Sequoias. Natl. Parks Conserv. Mag. 44, 16–22.
Kilgore, B.M., and Taylor, D. (1979). Fire history of a sequoia-mixed conifer forest. Ecology 60, 129–142.
Meyer, M.D.; Long, J.W.; Safford, H.D., eds. 2021. Postfire restoration framework for national forests in California. Gen. Tech. Rep. PSW-GTR-270. Albany, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station. 204 p.
Miller, J.D., and Thode, A.E. (2007). Quantifying burn severity in a heterogeneous landscape with a relative version of the delta Normalized Burn Ratio (dNBR). Remote Sensing of the Environment 109, 66–80.
Miller, J.D., Knapp, E.E., Key, C.H., Skinner, C.N., Isbell, C.J., Creasy, R.M., and Sherlock, J.W. (2009). Calibration and validation of the relative differenced Normalized Burn Ratio (RdNBR) to three measures of fire severity in the Sierra Nevada and Klamath Mountains, California, USA. Remote Sensing of the Environment. 113, 645–656.
National Park Service. 2012. NPS Policy Memorandum 12-02 Applying National Park Service Management Policies in the Context of Climate Change.Parks, S.A., and Abatzoglou, J.T. (2020). Warmer and Drier Fire Seasons Contribute to Increases in Area Burned at High Severity in Western US Forests From 1985 to 2017. Geophys. Res. Lett. 47, e2020GL089858.
Parsons, D.J., and DeBenedetti, S.H. (1979). Impact of fire suppression on a mixed-conifer forest. For. Ecol. Manag. 2, 21–33.
Shive, K., Weunschel, A., Hardlund, L., Meyer, M.D., and Morris, S. Declining fire resilience in the long-lived and fire-dependent Sequoiadendron giganteum (giant sequoia). J. For. Ecol. Manag. in review.
Sillett, S.C., Van Pelt, R., Carroll, A.L., Kramer, R.D., Ambrose, A.R., and Trask, D. (2015). How do tree structure and old age affect growth potential of California redwoods? Ecol. Monogr. 85, 181–212.
Stephens, S.L., Lydersen, J.M., Collins, B.M., Fry, D.L., and Meyer, M.D. (2015). Historical and current landscape-scale ponderosa pine and mixed conifer forest structure in the Southern Sierra Nevada. Ecosphere 6.
Stephenson, N.L. (1994). Long-term dynamics of giant sequoia populations: implications for managing a pioneer species. In Proceedings of the Symposium on Giant Sequoias: Their Place in the Ecosystem and Society, 23–25 June 1992, Visalia, California, USA, P.S. Aune, ed. (Visalia, California, USA: U.S. Forest Service), pp. 56–63.
Stephenson, N.L. 1996. Ecology and Management of Giant Sequoia Groves in Sierra Nevada Ecosystem Project Final Report to Congress. vol II: Assessments and Scientific Basis for Management Options. Davis: University of California, Centers for Water and Wildland Resources.
Stephenson, N.L., and Brigham, C. (2021). Preliminary Estimates of Sequoia Mortality in the 2020 Castle Fire (Sequoia and Kings Canyon National Parks: National Park Service).
Stohlgren, T. J. 1991. Size distributions and spatial patterns of giant sequoia (Sequoiadendron giganteum) in Sequoia and Kings Canyon National Parks, California. Technical Report No. 43, Cooperative National Park Resources Study Unit, University of California, Davis, CA. USDI National Park Service.
Swetnam, T.W. (1993). Fire History and Climate Change in Giant Sequoia Groves. Science 262, 885–889.
Swetnam, T.W., Baisan, C.H., Caprio, A.C., Brown, P.M., Touchan, R., Anderson, R.S., and Hallett, D.J. (2009). Multi-Millennial Fire History of the Giant Forest, Sequoia National Park, California, USA. Fire Ecol. 5, 120–150.
van Wagtendonk, K. (2012). Fires in Previously Burned Areas: Fire Severity and Vegetation Interactions in Yosemite National Park. Rethink. Prot. Areas Chang. World Proc. 2011 George Wright Soc. Bienn. Conf. Parks Prot. Areas Cult. Sites.
Wayman, R.B., and Safford, H.D. (2021). Recent bark beetle outbreaks influence wildfire severity in mixed-conifer forests of the Sierra Nevada, California, USA. Ecol. Appl. 31, e02287.
Williams, A.P., Abatzoglou, J.T., Gershunov, A., Guzman-Morales, J., Bishop, D.A., Balch, J.K., and Lettenmaier, D.P. (2019). Observed Impacts of Anthropogenic Climate Change on Wildfire in California. Earths Future 7, 892–910.
Young, D.J.N., Stevens, J.T., Earles, J.M., Moore, J., Ellis, A., Jirka, A.L., and Latimer, A.M. (2017). Long-term climate and competition explain forest mortality patterns under extreme drought. Ecol. Lett. 20, 78–86.