Effects of recent wildfires on giant sequoia groves were anomalous at millennial timescales: a response to Hanson et al.

Stephenson, Nathan L.; Soderberg, David N.; Flickinger, Joshua A.; Caprio, Anthony C.; Das, Adrian J.

doi:10.1186/s42408-024-00316-5

Forum
Open access
Published: 30 September 2024

Effects of recent wildfires on giant sequoia groves were anomalous at millennial timescales: a response to Hanson et al.

Nathan L. Stephenson ORCID: orcid.org/0000-0003-0208-7229¹,
David N. Soderberg¹,
Joshua A. Flickinger²,
Anthony C. Caprio² &
…
Adrian J. Das¹

Fire Ecology volume 20, Article number: 89 (2024) Cite this article

540 Accesses
1 Altmetric
Metrics details

Abstract

Background

The giant sequoia (Sequoiadendron giganteum [Lindley] Buchholz) of California’s Sierra Nevada recently suffered historically unprecedented wildfires that killed an estimated 13–19% of seed-bearing sequoias across their native range. Hanson et al. recently sought to characterize post-fire reproduction in two severely burned sequoia groves, but their two papers (1) inaccurately portrayed sequoia fire ecology, (2) had methodological flaws, and (3) without supporting evidence, questioned efforts to prevent large, stand-replacing wildfires and to plant sequoia seedlings in areas of low post-fire regeneration.

Results

Our analyses and literature review contradict many of Hanson et al.’s claims and implications. First, evidence indicates that preceding the recent wildfires, large, contiguous areas (>10 to >100 ha) of fire severe enough to kill most sequoias had been absent for at least a millennium, and probably much longer. The ancient sequoia fire regime was instead overwhelmingly dominated by surface fires in which most forest area burned at low or moderate severity interspersed with small forest gaps (hundredths of a hectare to a few hectares) created by local patches of higher-severity fire, within which most mature sequoias survived and most successful reproduction occurred. Prescribed fires have typically mimicked ancient fires and induced adequate sequoia regeneration. In contrast, in some extensive areas where recent wildfires killed most (or all) mature sequoias, regeneration has been well below historical levels, threatening a net loss of sequoia grove area. Methodologically, Hanson et al. reported sixfold greater post-fire sequoia seedling densities than others who sampled the same area; our assessments suggest their higher densities may have largely resulted from plot-placement bias. Finally, Hanson et al.’s comparisons of median seedling densities were inappropriate.

Conclusions

Hanson et al. questioned efforts to prevent large, high-severity wildfires in sequoia groves but did not acknowledge (1) that past fires sustained sequoia reproduction without the deaths of large fractions of mature sequoias, (2) the anomalous effects of recent wildfires, and (3) the acute conservation threat of losing large fractions of seed-bearing sequoias. Hanson et al.’s further implication, made without supporting evidence, that decisions to plant sequoia seedlings may be unwarranted ignores research showing that recent post-wildfire regeneration has often been well below historical levels.

Resumen

Antecedentes

Las sequoias gigantes (Sequoiadendron giganteum [Lindley] Buchholz) de la Sierra Nevada de California sufrieron incendios históricos sin precedentes que mataron aproximadamente un 13-19% de las sequoias portadoras de semillas a lo largo de su hábitat nativo. Hanson et al. buscaron recientemente caracterizar la reproducción post fuego en dos rodales de sequoias severamente quemados, aunque sus dos trabajos (1) describen inadecuadamente la ecología del fuego de las sequoias, (2) contienen problemas metodológicos, y (3) sin fundamentos, cuestionan los esfuerzos para prevenir grandes incendios que reemplazan rodales y que proponen plantar plántulas de sequoia en áreas con baja regeneración post-fuego.

Resultados

Nuestros análisis y la revisión bibliográfica contradicen mucho de lo que reivindican y las implicancias que sugieren los trabajos de Hanson et al. En primer lugar, la evidencia indica que, en áreas contiguas (>10 a >100 ha), fuegos recientes capaces de matar a las sequoias habían estado ausentes por al menos un milenio, y probablemente por más tiempo. El antiguo régimen de fuego de las sequoias estaba abrumadoramente dominado por fuegos de superficie, los cuales quemaban las áreas de bosque de manera superficial a baja o moderada severidad entremezclando áreas quemadas con pequeños parches de bosques sin quemar (desde cientos de ha a pocas ha), creadas por parches locales quemados a alta severidad dentro de los cuales la mayoría de las sequoias sobrevivieron y ocurrió una reproducción exitosa. Las quemas prescriptas han prácticamente copiado los incendios antiguos e inducido una regeneración adecuada de sequoias. En contraste, en algunas áreas extensas donde fuegos recientes mataron la mayoría (o todas) las sequoias maduras, la regeneración ha estado bien por debajo de sus niveles históricos, amenazando con una pérdida neta de los rodales de sequoias. Metodológicamente, Hanson et al. reportaron densidades de sequoias post fuego que representan valores seis veces más altos que otros que muestrearon la misma área; nuestras determinaciones sugieren que sus reportes de densidades más altas pueden haber sido el resultado de un sesgo en la ubicación de las parcelas de medición. Finalmente, las comparaciones de las medianas de densidad de plántulas fueron inapropiadas.

Conclusiones

Hanson et al. cuestionaron los esfuerzos para prevenir fuegos grandes y severos en rodales de sequoias, pero fallaron en reconocer que: (1) los fuegos pasados mantuvieron la reproducción de las sequoias sin que ocurra la muerte de grandes fracciones de rodales de sequoias maduras: (2) los efectos anómalos de fuegos recientes, (3) la aguda amenaza a la conservación por la pérdida de grandes fracciones de árboles semilleros. Las implicancias derivadas por Hanson et al., hechas sin las evidencias de soporte y que sostienen que el plantar plántulas de sequoia puede no justificarse, ignoran investigaciones que muestran que la reciente regeneración de sequoias post fuego ha estado muy por debajo de sus niveles históricos.

Introduction

The world’s most massive tree—the giant sequoia (Sequoiadendron giganteum [Lindley] Buchholz) of California’s Sierra Nevada—occurs naturally in fewer than 100 populations (“groves”) that collectively span only ~10,300 ha. Despite the exceptional fire resistance of sequoias, historically severe wildfires have recently killed an estimated 13–19% of mature sequoias across their native range (Stephenson and Brigham 2021; Shive et al. 2021, 2022). Some areas of sequoia forest burned in historically unprecedented crown fires, which burned cones out of the sequoias’ crowns and killed much of the local seed source (Soderberg et al. 2024). At the same time, historically exceptional drought and warmth (Williams et al. 2022; Stephenson et al. 2024) almost certainly reduced germination and survival of post-fire sequoia seedlings (Harvey et al. 1980). Consequently, sequoia seedling densities following recent wildfires often were significantly—and sometimes dramatically—lower than historical reference densities (Soderberg et al. 2024; Stephenson et al. 2024). Where most (or all) seed-bearing sequoias were killed by the wildfires, inadequate post-fire reproduction ultimately could contribute to a net loss of grove area (Soderberg et al. 2024; Stephenson et al. 2024).

In two recent papers, Hanson et al. (2024a, 2024b) sought to characterize post-fire sequoia reproduction in two severely burned sequoia groves: Redwood Mountain Grove (Kings Canyon National Park, California, USA), which burned in the 2021 KNP Complex wildfire, and Nelder Grove (Sierra National Forest, California, USA), which burned in the 2017 Railroad wildfire. However, both papers contained three broad classes of problems. First, the papers inaccurately portrayed key aspects of giant sequoia fire ecology, which apparently contributed to erroneous framing and interpretations. Second, both papers had methodological problems. Finally, the preceding problems apparently contributed to several of the papers’ unsupported and incorrect claims and conclusions.

We address these problems in three sections. The first reviews some key aspects of the broad literature on giant sequoia fire ecology, demonstrates that the effects of recent wildfires were anomalous at millennial timescales, and briefly reviews the vulnerability of sequoia reproduction to subsequent wildfires and ongoing environmental changes. The second section systematically assesses possible causes of Hanson et al.’s (2024a) reported sixfold greater post-fire seedling densities than those reported by Soderberg et al. (2024) within the same burned areas, and demonstrates that Hanson et al.’s (2024a, 2024b) use of medians was inappropriate. Finally, we correct some of Hanson et al.’s unsupported and erroneous claims and conclusions regarding efforts to reduce the threat of anomalously severe wildfires in sequoia groves, and efforts to plant sequoia seedlings in areas with low natural post-fire reproduction.

Throughout, we particularly focus on the severely burned portions of Redwood Mountain Grove (the grove studied by Hanson et al. 2024a) because this grove (1) has a rich, 60-year history of foundational research on sequoia fire ecology, (2) had its post-fire reproduction independently assessed by Soderberg et al. (2024), thus allowing direct comparisons with Hanson et al.’s (2024a) results, and (3) is the world’s second-largest sequoia grove (after Mountain Home Grove), comprising 10% of all native grove area in existence. In our analyses and assessments, we define high-severity burn areas as those with remotely sensed relativized differenced normalized burn ratio (RdNBR) >640 (Miller and Thode 2007). In the recently burned Redwood Mountain and Nelder groves (the groves studied by Hanson et al. 2024a, 2024b), RdNBR values >640 were associated with ~90–100% mortality of large sequoias (Shive et al. 2022; Soderberg et al. 2024). Because the distinction is important for understanding post-fire regeneration (see the next section), within high-severity burn areas (i.e., RdNBR >640), we further distinguish between areas where most sequoias were killed by crown scorch (i.e., killed by the convective heat of high-intensity surface fire without foliage combustion) and areas where most sequoias were killed by crown fire (foliage combustion) (Fig. 1).

Throughout, we define mature sequoias as those ≥100-cm diameter at breast height (DBH), thus slightly modifying Harvey et al.’s (1980) definition of mature sequoias as those that had reached the 3.5 to 4.5 ft DBH class (i.e., ≥106.7 cm DBH). In turn, we define millennial sequoias as those ≥320 cm DBH, which is a diameter that optimally distinguishes sequoias ≥1000 years old from those <1000 years old (see the next section and the Supplementary Information).

Giant sequoia fire ecology

The nature of past fire regimes and effects

Although the literature is far too extensive to fully review here, our knowledge of past fire regimes and their effects on sequoia groves and regeneration is based on many dozens of studies spanning six decades and using a broad array of complementary approaches: fire history from tree rings; demographic analysis; old plot data; old written accounts; repeat photography; age structure analysis; size structure analysis; physical legacies of past forest conditions; analyses of subfossil pollen, charcoal, and macrofossils; biological inference; and contemporary analogs (see the reviews in Stephenson 1996, 1999; Stephenson et al. 2024). These studies have been remarkably consistent in their alignment with the following summary (also see Kilgore and Taylor 1979; Swetnam 1993; Stephenson 1994, 1996, 1999; Stephenson et al. 1991, 2024; Swetnam et al. 2009).

The extensive, multi-millennial tree-ring record of fires in sequoia groves (Fig. 2) shows that the pre-fire-exclusion fire regime of sequoia groves was dominated by frequent surface fires. On average, at least one fire burned somewhere within a large sequoia grove every 2 years, and a fire burned at the base of any individual sequoia every 15 to 17 years (Kilgore and Taylor 1979; Swetnam et al. 2009). Thus, a typical 1000-year-old sequoia will have lived through more than 60 fires, and a 3000-year-old sequoia may have lived through as many as 200 fires. The fires were characterized by a large forest matrix that burned at low or moderate severity (where the forest canopy remained mostly intact), interspersed with small forest gaps created by local patches of higher-severity fire. The gaps typically ranged in size from hundredths of a hectare up to a few hectares, with a probable modal gap size on the order of ~0.1 ha (Stephenson 1996, 1999).

Although occasional individual trees—usually smaller trees—would sometimes torch (have their crowns combust), the small fire-caused gaps were largely created by treefalls and especially by crown scorch from the convective heat of surface fires that locally killed many (or most) pines, firs, incense cedars, and oaks. That is, the gaps were rarely (if ever) created by “patches of crown fire” (foliage combustion), as was incorrectly stated by Hanson et al. (2024a). Importantly, large sequoias within the gaps usually survived (Fig. 1A). Most sequoia seedling establishment that went on to successfully produce mature sequoias occurred in these gaps (Stephenson 1996). The past fire regime thus ultimately created a fine-scale mosaic of forest patches of different ages (Fig. 3) (Bonnicksen and Stone 1982, Stephenson 1996).

Age structure analysis coupled with demographic analysis has shown that the past surface fire regime and its associated sequoia regeneration sustained stable sequoia populations over at least the last two millennia (York et al. 2013). Specifically, frequent fire-induced sequoia recruitment maintained sequoia populations with age structures closely approximating stationary age distributions (York et al. 2013).

Sequoia reproduction first increases, then declines with increasing fire severity

Stephenson et al. (2024) offered a detailed review of the relationship between fire severity and sequoia reproduction. Here we briefly summarize the current state of knowledge.

Before the historically unprecedented wildfires of 2015, 2017, 2020, and 2021 (Shive et al. 2021, 2022; Stephenson and Brigham 2021), researchers reported a positive relationship between local fire severity and sequoia reproduction (see the reviews in Stephenson 1996; Stephenson et al. 2024). Although several factors contributed to this relationship, an especially critical factor was that local patches of higher-intensity surface fire would often scorch portions of sequoia crowns—causing the scorched cones to open and release their abundant seeds—while nearly always leaving the tree alive (Harvey et al. 1980) (Fig. 1A). In contrast, some areas of recent wildfires have burned at historically unprecedented severity (often with RdNBR >1000; Soderberg et al. (2024)), killing most (or all) sequoias in patches sometimes exceeding 100 ha. Where the sequoias in these patches died from crown scorch (i.e., if they were killed by convective heat from surface fire), their cones remained attached in their crowns and, as in the past, the cones were likely to release abundant seeds (Fig. 1B). In contrast, in those areas where sequoias died from crown fire (foliage combustion), most cones were burned out of the sequoias’ crowns and fell into the surface fire below (Soderberg et al. 2024), where they either combusted or may have been heated to temperatures that killed many of their seeds (Fig. 1C).

Thus, grove areas where most sequoias were killed in recent crown fires have had lower seedling densities than areas where most sequoias were killed by high-severity surface fires (Soderberg et al. 2024). That is, we now know that the relationship between fire severity and sequoia seedling density is hump-shaped, peaking at the lower levels of high severity that were typical of the past and then declining at the new extremes of high severity that include crown fire (Meyer et al. 2024; Soderberg et al. 2024). Similarly, other Sierra Nevada tree species show declining seedling densities at the extremes of high-severity fire (Welch et al. 2016).

Although Hanson et al. (2024b) reported that they did not detect the hump-shaped relationship between fire severity and sequoia seedling density in their Nelder Grove data, their small plots (0.008 ha) contributed to high among-plot variance (see their Fig. 2), and their overall sampling effort (23 plots collectively spanning 0.18 ha in a single grove) was an order of magnitude less than each of the two recent studies that found declining seedling densities at the new extremes of high-severity fire. Specifically, Soderberg et al. (2024) used 116 plots, collectively spanning 5.9 ha, to sample four groves that burned in the 2021 KNP Complex wildfire, and Meyer et al. (2024) used 113 plots, collectively spanning 5.65 ha, to sample seven groves that burned in the 2021 Windy wildfire.

Crown fires have been absent for at least a millennium

In their opening paragraphs, Hanson et al. (2024a, 2024b) seemed to equate giant sequoia with other coniferous tree species that are famously adapted to crown fire, and they suggested that optimal conditions for giant sequoia reproduction are “created in particular by patches of crown fire” (Hanson et al. 2024a). The authors offered no evidence supporting these claims. Here we present evidence that, until the recent wildfires, giant sequoia populations had not experienced significant crown fire—or any fire severe enough to kill most sequoias over large areas—for at least the preceding millennium.

To illustrate the long-term absence of large crown fires, we examined the abundances and distributions of sequoias ≥1000 years old (“millennial sequoias”). Because sequoias that have lost all their foliage do not resprout (Schubert and Beetham 1962; Stephens and Finney 2002; Shive et al. 2022), the frequent fires of the past that millennial sequoias lived through (Kilgore and Taylor 1979; Swetnam 1993; Swetnam et al. 2009) were clearly surface fires, not crown fires. That is, the presence of a living millennial sequoia indicates the absence of crown fire—or any fire severe enough to completely (or nearly completely) scorch the sequoia’s crown—at that point for at least the preceding 1000 years.

Based on data from 458 cut sequoia stumps whose ages were determined by Huntington (1914), we distinguished living millennial sequoias as those with trunk diameters at breast height (DBH) ≥320 cm (methods are in the Supplementary Information; data are summarized in Tables S1 and S2). This is an optimal trunk diameter that yields comparable estimated proportions of false-positive (17%) and false-negative (15%) classifications (Table S2). Similarly, old cut sequoia stumps ≥259 cm DBH were considered to be millennial sequoias that would still be alive today if they had never been cut; the smaller stump diameter compared to living sequoias accounts for the stumps’ lost diameter growth, lost bark, and lost sapwood (see the Supplementary Information).

Millennial sequoias were well-distributed throughout the 834-ha stem-mapped portion of Redwood Mountain Grove (where Hanson et al. (2024a) conducted their work), with only a few small, scattered patches (<10 ha) lacking them (Fig. 4). That is, the stem-mapped portion of the grove—which represents 8% of all native sequoia grove area in existence—had not experienced a large, stand-replacing fire for at least 1000 years preceding the KNP Complex wildfire. Using similar lines of evidence, Kilgore and Taylor (1979) concluded that the grove had not experienced a crown fire of any meaningful size for at least the preceding 2000 years.

Furthermore, it is highly unlikely that even the small, scattered patches in Redwood Mountain Grove that lack millennial sequoias were caused by small, local crown fires. First, some of the patches may represent areas of the recent grove expansion that was reported by Harvey et al. (1980), perhaps related to cooler conditions during the Little Ice Age (ca. 1300 to 1850 CE). Second, some of the patches are likely a consequence of substrate limitations, such as meadows, rock outcrops, or steep, wet slopes. Third, groups of sequoias can be killed by disturbances other than fire, most notably by avalanches, landslides, and wind (see the Supplementary Information). Finally, even if a few of the small patches reflect areas where now-absent millennial sequoias were killed by single fires or a series of fires sometime over the last several centuries, several lines of evidence show that the deaths would almost certainly have been by crown scorch, basal girdling, or structural collapse caused by surface fires, not crown fires (e.g., Kilgore and Taylor 1979).

Millennial sequoias are similarly ubiquitous in other sequoia groves. Within Sequoia and Kings Canyon national parks, all 30 censused and stem-mapped sequoia groves >1 ha contained at least moderate densities of millennial sequoias (Table 1; see the Supplementary Material for methods), with millennial sequoias scattered broadly within their boundaries. Of the 22 groves >10 ha (i.e., for those groves at least as large as the unusually large high-severity burn areas of recent wildfires), even the grove with the lowest density of millennial sequoias—the 88.5-ha Castle Creek Grove—had 88 millennial sequoias distributed within its boundary, with no extensive areas devoid of millennial sequoias (Fig. S1). Thus, while a more in-depth analysis is beyond the scope of this paper, within the cumulative 3729 ha of censused and stem-mapped sequoia groves in Sequoia and Kings Canyon national parks (Table 1)—representing 36% of all grove area across the species’ native range—the available data are consistent with the absence of extensive, stand-replacing fires for at least the millennium preceding the recent wildfires (e.g., Table 1, Fig. 4, Fig. S1; also see the next subsection).

Table 1 Estimated numbers and densities of millennial sequoias (those ≥1000 years old when censused in the 1960s and 1970s) in all censused and stem-mapped grove areas >1 ha in Sequoia and Kings Canyon national parks, California, USA

Full size table

Beyond the censused groves, the only large, contiguous grove areas dominated by young, even-aged sequoias (an indicator of stand-replacing disturbance) of which we are aware fall within the ~23% of native grove area that was logged of most of its sequoias during the last ~150 years (Stephenson 1996). Furthermore, after examining dozens of sequoia groves Hartesveldt (1964) reported that standing sequoia snags that were potentially killed by fire “are rarely found in groups of more than two or three,” and that this “[s]carcity of clusters indicates few fires of any intensity” (where, from context, “any intensity” appears to mean “severe enough to kill groups of large sequoias”).

The multi-millennial sequoia tree-ring record (e.g., Swetnam 1993; Caprio et al. 1994; Swetnam et al. 2009) is consistent with the preceding observations. Even the most severe fire found in the sequoia tree-ring record—the 1297 CE fire in Mountain Home Grove (Caprio et al. 1994)—shows no evidence of including large patches high-severity fire (using contemporary definitions of high severity; see below). Millennial sequoias survived and recorded the fire in their rings across all sampled areas of the grove, and a large majority of samples (89%) showed either (1) little change in post-fire growth, suggesting local low-severity fire (29%), or (2) often substantial increases in post-fire growth (60%), suggesting the death of many competing pines, firs, and incense-cedars but <50% crown scorch to the sequoias themselves (Caprio et al. 1994). The remaining 11% of sampled sequoias showed initial post-fire growth declines—typically lasting about 20 years and suggesting >50% crown scorch—but even those sequoias often grew near other sequoias that showed little change in, or increases in, post-fire growth, suggesting that the areas of >50% crown scorch may have been localized rather than extensive. Although we will never know the proportion of large sequoias killed in the 1297 fire (their snags and logs have long since burned), the nature of pre-fire-exclusion fuels (see below, and Shive et al. 2022) means we can reasonably surmise that most (or all) sequoia deaths were likely from crown scorch, not crown fire. (For readers of Caprio et al. (1994), it is important to note that the authors’ definition of “high severity” fire was notably less severe than contemporary definitions. Specifically, current definitions of high-severity fire [i.e., RdNBR >640] are associated with 75–100% mortality of large sequoias (Shive et al. 2022; and see below). In sharp contrast, Caprio et al. defined “high severity” as being associated with sequoias that likely suffered <50% crown scorch and showed substantially increased growth after the fire and defined “very high severity” to include sequoias with >50% crown scorch that showed initial growth declines but did not die, instead recovering and later showing growth increases.)

Finally, our evidence that meaningful crown fire was absent from sequoia groves also is not contradicted by the size, behavior, and effects of the 1987 Pierce wildfire in the Redwood Mountain Grove, mentioned by Hanson et al. (2024a). Until the 2015 Rough wildfire (Shive et al. 2022), the Pierce wildfire was the most severe modern fire recorded in a sequoia grove. The fire killed 14 mature sequoias (Stephenson et al. 1991)—two of which were millennial sequoias—after burning “approximately 20 ha of the grove, primarily at high severity (>75% basal area mortality one year after fire …)” (Meyer and Safford 2011). Although several individual trees torched (had their crowns combust) in the wildfire, most trees did not; the fire was predominantly a surface fire. Even then, some of the fire’s behavior and effects were largely outside the normal bounds of ancient fires in sequoia groves. The area that burned at high severity had been logged of many of its sequoias before 1915 (Meyer and Safford 2011; see the mapped stumps on the west side of Redwood Mountain Grove in Fig. 4). The Pierce fire thus burned through post-logging sequoia woody debris (which decomposes slowly and remains flammable), dense patches of shrubs, and areas of dense post-logging sequoia regeneration with crowns that still extended to ground level which, due to active fire exclusion, had not been thinned by subsequent surface fires. That is, the hazardous fuel profile present during the 1987 wildfire was largely an unnatural product of logging followed by a century of fire exclusion and included dense ladder fuels that burned at high intensity, conducting fires into the crowns of some larger trees. In sharp contrast, when the Pierce wildfire reached an area that had recently been prescribed burned, it quickly died down and was easily contained (Nichols 1989; Stephenson et al. 1991; Stephenson 1996). Like prescribed fires (Fig. 5), the frequent fires of the past greatly reduced ladder fuels and kept surface fuel loads low in sequoia groves (Kilgore and Sando 1975, Keifer 1998, Kilgore and Taylor 1979, Shive et al. 2022; also see Keifer et al. 2006, van Mantgem et al. 2016).

Moving beyond the giant sequoia—mixed conifer forest (a moist sub-type of the broader Sierra Nevada mixed conifer forest (Rundel 1972; Halpin 1995; Fites-Kaufman et al. 2007)), Hanson et al. (2024b) cited two papers (Baker 2014; Baker and Hanson 2017) concluding that within Sierra Nevada mixed conifer forests, high-severity fires “hundreds of hectares in size” occurred at rotation intervals “typically spaced by three to four centuries on average.” But if large, high-severity fires comparable to recent wildfires had rotation intervals of 300–400 years in sequoia groves (as Hanson et al. seemed to suggest), contemporary groves would be much more heavily dominated by young sequoias—with most grove areas resembling forests at various stages of recovery from logging—and would contain few, if any, millennial sequoias (see the next subsection). Regardless, although the two papers Hanson et al. cited focused on non-sequoia forest types, we note that several other studies using a variety of approaches have concluded that the papers had methodological problems and that large, severe fires were almost certainly less frequent in Sierra Nevada mixed conifer forests than the papers would suggest (e.g., see the reviews and summaries in Levine et al. 2017, Hagmann et al. 2018, Levine et al. 2019, Hagmann et al. 2021, and Prichard et al. 2021).

Recent high-severity wildfires were anomalous at millennial timescales

The preceding subsection refutes Hanson et al.’s (2024a, 2024b) apparent suggestion that crown fires—or any fires severe enough to kill most sequoias over large areas—were a normal part of giant sequoia fire regimes of the past. Now we describe aspects of the anomalous extent and effects of recent large patches of high-severity fire—that is, patches larger than ~10 ha in which most or all sequoias were killed, whether by crown fire or by crown scorch (heating from intense surface fire). Others have already discussed the anomalous extent of high-severity patches relative to the preceding ~100 years (e.g., Shive et al. 2022); we thus briefly assess the anomalous nature of the patches at millennial timescales.

In the preceding subsection we showed that the world’s second-largest sequoia grove, Redwood Mountain Grove, had not experienced even a modestly sized stand-replacing fire over at least the millennium preceding the 2021 KNP Complex wildfire (Fig. 4). In sharp contrast, 257 ha (31%) of the 834 ha of stem-mapped grove area burned at high severity in the KNP Complex wildfire (Fig. 4). Soderberg et al.’s post-fire sequoia tree census plots (see Soderberg et al. (2024) and the Supplementary Information) collectively sampled 14% of these high-severity areas, revealing that 31% of mature sequoias (those ≥100 cm DBH) had experienced 100% crown combustion and more than half had experienced ≥85% crown combustion, indicating relatively extensive areas of crown fire. Soderberg et al.’s data further showed that 89% of sampled mature sequoias were dead (0% live crown) 1 year after the fire. An additional 8% of mature sequoias had ≤10% of their live crowns remaining (averaging <6% live crown), meaning they had a high probability of dying within the next few years (Shive et al. 2022). It is thus reasonable to assume that up to 97% of seed-bearing sequoias in the severely burned parts of the grove died in the fire or soon thereafter. For millennial sequoias (those ≥320 cm DBH), the corresponding numbers were 92% dead 1 year post-fire and more than 6% with a high probability of dying (i.e., with ≤10% of their live crowns remaining), potentially totaling up to 98% mortality within a few years of the fire.

The largest single area of contiguous high-severity fire in Redwood Mountain Grove (137 ha; the central southern red area in Fig. 4) contained 281 millennial sequoias before the KNP Complex wildfire. From the preceding paragraph, 92–98% of these sequoias were likely killed by the wildfire, leaving only 6 to 22 surviving millennial sequoias. A further eight millennial sequoias fell within the ~4 ha of moderate-severity burn patches inside the boundaries of the contiguous high-severity burn area (Fig. 4). If we assume these eight sequoias suffered 0–45% mortality (e.g., Shive et al. 2022), four to eight of them survived the wildfire, for a total of ~10 to 30 millennial sequoias surviving within the entire 141-ha area. Before the wildfire, the density of millennial sequoias within the 141-ha area was 2.05 ha⁻¹, nearly identical to the 1.98 ha⁻¹ average density of millennial sequoias across the 22 censused groves that were >10 ha (Table 1). In sharp contrast, the estimated 0.07–0.21 ha⁻¹ density of surviving millennial sequoias after the wildfire is at least an order of magnitude less than the pre-fire density, and ranges from ~fivefold to ~70-fold lower than any of the 22 individual censused groves that were >10 ha (Table 1). Such markedly low post-fire densities of living millennial sequoias across 141 ha—an area larger than most individual sequoia groves—is clearly anomalous (Table 1) and, based on available data, without precedent during at least the preceding millennium.

Like millennial sequoias, the overall densities of mature sequoias (those ≥100 cm DBH) within the contiguous 141-ha area of Redwood Mountain Grove were likely reduced by 89–97% (see the preceding paragraphs). That is, for the first time in at least the last millennium, the world’s second-largest sequoia grove now has some large, contiguous areas containing few living, seed-bearing sequoias, and large areas that are distant from surviving seed trees; for example, 43% of Soderberg et al.’s (2024) randomly located seedling plots were >250 m from a surviving, seed-bearing sequoia. More broadly, more than 13% of sequoia grove area across the species’ native range burned at high severity during the summers of 2020 and 2021, much of it in large contiguous areas (Shive et al. 2021; Stephenson and Brigham 2021). That is, the anomalous conditions now found in parts of Redwood Mountain Grove are also found in many other groves across the species’ range (e.g., Shive et al. 2022).

Fuels accumulate rapidly after high-severity wildfires (e.g., Lydersen et al. 2019), and if they are not reduced, future wildfires could kill most or all seedlings that might have become established in those high-severity areas (see the subsection “Distance to surviving seed trees reflects vulnerability to future loss”). But surviving seed trees may be too distant to supply a sufficient new cohort of seedlings—a phenomenon documented in other coniferous forests of western North America (e.g., Shive et al. 2018; Stevens-Rumann and Morgan 2019; Coop et al. 2020; Stewart et al. 2021; Davis et al. 2023). Thus, in the face of future wildfires Redwood Mountain Grove is now vulnerable to as much as a ~30% net loss of area, with some other groves vulnerable to even greater losses (Shive et al. 2021; Stephenson and Brigham 2021).

The KNP Complex wildfire first entered Redwood Mountain Grove as a headfire moving up slope. Much of the grove area that the fire first encountered had never been prescribed burned, and these were the areas that burned mostly at high severity (Fig. 4). But when the fire then encountered areas that had been prescribed burned within the last ~10–15 years, it largely became a low- to moderate-severity surface fire, leaving most mature sequoias alive. That is, in those areas where prescribed fires had restored both surface fuels and ladder fuels to conditions similar to the pre-fire-exclusion past (e.g., Fig. 5; Kilgore and Sando 1975; Keifer 1998; Keifer et al. 2006; van Mantgem et al. 2016), fire behavior and effects also became more similar to those of the past.

Giant sequoia is maladapted to crown fire

The preceding subsections presented evidence that large areas of crown fire (or any fire severe enough to kill most sequoias) have been absent from sequoia groves for at least the millennium preceding recent wildfires. We now greatly expand this time horizon, summarizing evidence that crown fire has almost certainly not been a significant factor affecting giant sequoias over evolutionary timescales.

Some of the typical traits of coniferous tree species that are adapted to crown fire include thin bark, relatively short stature, retention of dead lower branches, and short time to maturity (Keeley and Zedler 1998; Schwilk and Ackerly 2001; Stevens et al. 2020). Giant sequoia shows diametrically opposite traits (i.e., it has fire resistance traits): thick bark (often >20 cm), great height (it is among the tallest 0.01% of tree species on Earth), non-retention of dead lower branches (height to first branches, living or dead, often exceeds 20 m), and a relatively long time to maturity (Schubert and Beetham 1962; Hartesveldt et al. 1975; Harvey et al. 1980; Sillett et al. 2015; Stevens et al. 2020). But most important to the current discussion, giant sequoia reproduction is clearly maladapted to crown fire.

Sequoias reproduce only by seed, from cones that hang from long, thin peduncles (Fig. 6A) (Buchholz 1938). As alluded to in preceding subsections, during the typical surface fires of the past the convective heat pulse from local patches of high-intensity surface fire would often scorch sequoias’ lower crowns (heating, but not combusting the foliage, cones, and peduncles supporting cones) (Fig. 1A), ultimately causing the cones—which remained attached high in the tree—to dry and release their seeds (Harvey et al. 1980). In contrast, during the recent, anomalous crown fires the long, thin peduncles that support sequoia cones have burned (Fig. 1C), dropping most cones into the fire below (Soderberg et al. 2024) and killing many or most of their seeds.

In contrast to giant sequoia, coniferous species that evolved with crown fire usually have either sessile cones (adhering directly to woody branches or trunks), or very short, thick peduncles that resist combustion (Warren and Fordham 1978; Martín-Sanz et al. 2017; Reilly et al. 2019; Greene et al. 2024). Thus, unlike giant sequoia, most of their cones remain firmly attached to the tree during and after crown fires (Fig. 6B–D), allowing the seeds to survive the passage of the crown fire (Lamont et al. 1991). The cones subsequently dry and release their seeds into the post-fire environment.

Hanson et al.’s (2024a, 2024b) apparent contention that crown fires have been a significant part of giant sequoia fire regimes may have arisen from their conflation of two different definitions of serotiny. Giant sequoia is only serotinous in the broadest definition of the term: having prolonged canopy seed storage (Lamont et al. 1991). But giant sequoia is perhaps best called necriscent—seeds are released upon the death of cones by any cause, including but not limited to heat from surface fires (Harvey et al. 1980; Lamont 1991; Lamont et al. 1991). Yet in the introductions to their papers Hanson et al. (2024a, 2024b) conflated sequoia’s necriscence with pyriscence. The term pyriscent—the narrowest definition of serotiny—refers to species that require fire for seed release, usually because (in contrast to giant sequoia) the scales of pyriscent cones are held tightly closed by resins until the resins are melted by heat from a fire. In sharp contrast to the pyriscent species Hanson et al. (2024a, 2024b) seemingly implied were the serotinous equivalents of giant sequoia, sequoia offers a quintessential example of a necriscent species that is maladapted to crown fire, suggesting that over evolutionary time scales crown fire has not been present as a significant selective force on sequoias.

Distance to surviving seed trees reflects vulnerability to future loss

Hanson et al. (2024a, 2024b) found no relationship between sequoia seedling density and distance to the nearest seed tree that survived the wildfires in Redwood Mountain Grove and Nelder Grove, but apparently misinterpreted the significance of their findings. Specifically, Hanson et al. appeared to imply that the surviving seed trees were the seed source for their seedlings, even at distances >700 m. But this is not the case. Most sequoia seed dispersal is quite local (Harvey et al. 1980; Demetry 1995; Clark et al. 1999), and the seed source for most seedlings is not the distant surviving seed trees, but the nearby sequoias that were killed by the wildfire.

However, distance to surviving seed tree does reflect vulnerability to net loss of grove area during future wildfires. Fuels for future wildfires will include abundant logs and snags (from the trees killed by the initial high-severity fire) and shrubs, whose germination from a persistent soil seed bank was triggered by the fires (Collins and Roller 2013; Coppoletta et al. 2016; Lydersen et al. 2019; North et al. 2019). If these dense post-fire fuels are not reduced (such as by prescribed surface fires), areas that previously burned at high severity have a high probability of reburning at high severity, especially during extreme fire weather (Coppoletta et al. 2016; Harris and Taylor 2017; Lydersen et al. 2017, 2019; Taylor et al. 2021, 2022). If severely reburned areas are distant from surviving seed trees, a type conversion to shrub fields with few surviving tree seedlings, including few sequoia seedlings, would be possible and perhaps even probable (North et al. 2019; Stevens-Rumann and Morgan 2019; Coop et al. 2020; Stewart et al. 2021; Gill et al. 2022; Guiterman et al. 2022; Stevens-Rumann et al. 2022; Davis et al. 2023).

Despite Hanson et al.’s (2024b) apparent dismissal of this type-conversion threat and contention that “the young, naturally regenerating sequoia forests in high-severity fire patches are more likely to experience low-/moderate severity fire if a reburn occurs,” the risk of subsequent wildfires killing sequoia regeneration is substantial, as was vividly illustrated by the 2021 KNP Complex wildfire in Redwood Mountain Grove. The grove contained one of the largest concentrations of recent natural sequoia regeneration in the Sierra Nevada, in an area of a few hectares dominated by young sequoias and shrubs (Fig. 7A) (also see the descriptions and photos in Stephenson et al. 1991; Stephenson 1994). The sequoias had germinated in a large “hot spot” created by a prescribed fire in 1977, although all mature sequoias within the hot spot survived the initial fire (Stephenson et al. 1991; Stephenson 1994). In 2011, another prescribed fire was used to thin the young sequoias and reduce shrubs and surface fuels. The fire reduced the density of young sequoias by an estimated 25–40%, with fire managers recognizing that one or two additional prescribed fires would likely be required to create conditions resistant to a high intensity headfire. However, no additional prescribed fires had occurred by the time of the 2021 KNP Complex wildfire, when the young sequoias were 42 to 43 years old. Despite the substantial inherent resistance of young sequoias to moderate surface fires (Bellows et al. 2016; York et al. 2021), all the thousands of young sequoias in the patch perished in the wildfire, as did the mature seed trees that had survived the initial 1977 prescribed fire (Fig. 7B,C). Most of the young sequoias burned in crown fire.

Novel environmental conditions likely reduce seedling survival

As recently reviewed by Stephenson et al. (2024), several novel environmental conditions are likely to reduce survival of sequoia seedlings that germinated after recent wildfires. First, the exceptionally large forest gaps created by recent wildfires will typically retain less snow and will melt out earlier compared to the smaller gaps of the past (Stevens 2017; Gleason et al. 2019; Smoot and Gleason 2021; Hatchett et al. 2023), exacerbating the summer drought experienced by sequoia seedlings in large gaps. Second, areas of sequoia groves that experienced historically unprecedented crown fires often have a reduced (or absent) post-fire layer of leaf litter on the soil surface (cf. Welch et al. 2016), likely resulting in higher soil-surface temperatures, reduced soil moisture retention, and thus reduced sequoia seedling survival (Stark 1968). Third, both the direct effects of rising temperatures on sequoia seedlings (Stark 1968; Moran et al. 2019) and the indirect effects of rising temperatures on snowmelt timing and climatic water deficits (drought) (Andrews 2013; Harpold and Molotch 2015; Mote et al. 2018; Tercek et al. 2021) are likely to reduce sequoia seedling survival (Harvey et al. 1980; Mutch and Swetnam 1995). Finally, and especially in the face of increasingly extreme Sierra Nevada fire weather (Collins 2014; Williams et al. 2019; Goss et al. 2020; Gutierrez et al. 2021; Rother et al. 2022; Brown et al. 2023), those areas that previously burned at high severity have a high probability of reburning at high severity (Coppoletta et al. 2016; Harris and Taylor 2017; Lydersen et al. 2017, 2019; Taylor et al. 2021, 2022), thus killing larger proportions of sequoia seedlings than would have died in the past (Fig. 7). Each of these four novel environmental factors is discussed in greater detail in Stephenson et al. (2024).

The post-fire seedling reference densities reported by Stephenson et al. (2023, 2024) were based on data from a substantially more benign climatic period than that which prevailed during the critical first years following the recent wildfires in sequoia groves (Stephenson et al. 2024), and more benign than is expected for the future (e.g., Gonzalez 2012). Thus, Stephenson et al.’s (2023, 2024) reference densities are almost certainly conservative, underestimating the seedling densities needed to regenerate sequoia populations killed in high-severity areas of recent wildfires. (When applied to high-severity burn areas, Stephenson et al.’s reference densities are also inherently conservative because the data were predominantly from low- and moderate-severity burn areas; see Stephenson et al. 2024.) Yet Hanson et al. (2024a, 2024b) simply did not acknowledge the likely effects of novel environmental conditions on sequoia seedling survival.

Methodological problems

Assessing apparent data bias

Density differences between data sets

We would expect two independent groups, sampling the same high-severity burn areas in Redwood Mountain Grove during the same summer (2023), to find similar mean sequoia seedling densities—or, at a minimum, to find mean densities that were statistically indistinguishable. Yet Hanson et al.’s (2024a) arithmetic mean seedling density for high-severity burn areas (59,461 ha⁻¹) was more than sixfold greater than that found by Soderberg et al. (2024) (9455 ha⁻¹). As described in the Supplementary Information (also see Soderberg et al. 2024), we used a Bayesian negative binomial count model to describe uncertainty (as probability distributions) in mean estimated seedling densities for both data sets. Our analysis showed that the profound density difference is statistically highly significant, with a marginal probability of <0.0001 that Hanson et al.’s mean density is equal to or less than Soderberg et al.’s mean density (Fig. 8).

This strong statistical significance indicates that the difference between the means is not simply a consequence of Hanson et al.’s (2024a) weaker sampling effort relative to that of Soderberg et al. (2024) (0.136 ha vs. 1.92 ha cumulative sample areas, respectively, within the high-severity burn areas). Total numbers of plots were comparable between the data sets (see below), but the much smaller size of Hanson et al.’s plots almost certainly contributed to the eightfold greater density variance among Hanson et al.’s plots compared those of Soderberg et al. The greater uncertainty inherent in Hanson et al.’s data—due to smaller total sample area and higher among-plot variance—would be expected to reduce statistical power and weaken our ability to detect any possible difference in means between the two data sets. Yet we still detected a highly significant difference, overwhelmingly suggesting that the two data sets do not represent the same sample population.

Thus, because the striking density difference between the Hanson et al. (2024a) and Soderberg et al. (2024) data sets has a vanishingly small probability of being caused by chance alone, the most parsimonious explanation is that the difference is caused by bias in one—or both—of the data sets. Accordingly, in the following subsections we assess potential sources of bias in both Soderberg et al.’s (2024) and Hanson et al.’s (2024a) Redwood Mountain seedling density data.

Because all 46 of Soderberg et al.’s (2024) plots were in high-severity burn areas (RdNBR >640; Miller and Thode (2007)), all the following comparisons are with Hanson et al.’s (2024a) subset of plots that were likewise in high-severity burn areas. Soderberg et al.’s (2024) original analyses used MTBS RdNBR values, which are generally derived ~1 year post-fire. For consistent comparisons between data sets, we thus analyzed Hanson et al.’s (2024a) 43 plots that similarly had MTBS RdNBR values >640, rather than their 45 plots that had RAVG RdNBR values >640 (RAVG RdNBR values are generally derived ≤45 days after fire containment).

Seedling cohorts compared

Hanson et al. (2024a) seemed to suggest that they found greater seedling densities in Redwood Mountain Grove than Soderberg et al. (2024) because Hanson et al. censused seedlings 2 years post-fire, whereas Soderberg et al. (2024) purportedly censused seedlings 1 year post-fire, thus missing the second-year seedling cohort. However, Hanson et al. apparently overlooked the fact that Soderberg et al. (2024) censused seedlings in both the first and second years post-fire. That is, the observed sixfold difference in mean seedling densities is for seedlings measured by both groups in the second year post-fire and thus cannot be attributed to differences in the seedling cohorts compared.

Plot sizes

Relative to Hanson et al.’s (2024a) plots (1-m radius [70% of plots] and 2-, 5-, or 10-m radius [30% of plots]), Soderberg et al.’s (2024) Redwood Mountain plots were quite large (11.35-m radius [98% of plots] and 17.84-m radius [2% of plots]), raising the possibility that Soderberg et al.’s lower densities might be a consequence of systematic undercounting of seedlings in such large plots. Although undercounting is a real possibility in large plots (as is accidental double-counting of some seedlings), it defies reasonable expectation that careful, systematic censuses by an experienced, well-trained field crew would overlook >80% of seedlings, especially considering that, as is typical for post-fire sequoia seedling populations (Stephenson et al. 2024), the majority of seedlings present in 2023 belonged to the 2022 cohort and were thus 1 year old—large enough to be easily seen and counted (e.g., see Figs. 4 and 5 in Hanson et al. 2024a). Indeed, Hanson expressed concern to National Park Service managers about the high density of footprints left by the sampling crews throughout the Soderberg et al. plots he visited (C. M. Brigham, National Park Service, Three Rivers, California, USA, written communication, 2 August 2024), suggesting that the Soderberg et al. field crews were thorough in censusing all parts of the plots. Additionally, the seasonal timing of Soderberg et al.’s censuses (early in the summer, before reductions of seedling densities by summer mortality) suggests that, all else being equal, Soderberg et al. should have encountered denser seedlings than Hanson et al. (thus overcounting relative to Hanson et al.), who mostly sampled later in summer (see the Supplementary information).

Conversely, although the small size of Hanson et al.’s (2024a) plots likely contributed to the eightfold greater variance in their data compared to Soderberg et al.’s (2024) data, small plot size alone also cannot reasonably be expected to account for the substantial, highly significant mean density differences between the two data sets (see “Density differences between data sets”, above). But unlike large plots, small plots are particularly vulnerable to plot-placement bias (Elzinga et al. 1998), assessed next.

Plot-placement bias

It is highly unlikely that Soderberg et al.’s (2024) seedling densities were meaningfully affected by plot-placement bias. Plot centers were determined a priori using the Generalized Random Tessellation Stratified algorithm (GRTS), with an equal probability stratified sampling design providing a spatially balanced sample that has a true probability design, allowing valid inference for the entire study area (Stevens and Olsen 2004). In the field, none of Soderberg et al.’s a priori GRTS-determined plot locations was rejected or relocated. Plot centers were found using a mapping-grade GPS unit (Leica Zeno GG04 plus). Root mean square GPS errors for final plot centers averaged 0.52 m and ranged from 0.20 to 1.89 m—errors that were small (mean = 5%; range = 2 to 17%) relative to the 11.35-m or 17.84-m radii of the plots.

In contrast, we believe Hanson et al.’s (2024a) plots may have been vulnerable to significant plot-placement bias. We particularly focus on Hanson et al.’s 1-m radius plots, because the greatest disparity between Hanson et al.’s (2024a) and Soderberg et al.’s (2024) mean seedling densities is found in the 65% of plots with the greatest seedling densities, which in Hanson et al.’s case solely comprised 1-m radius plots (see the Supplementary Information). Hanson et al.’s 1-m radius plots would be particularly vulnerable to shifts in placement of even a few tens of centimeters; for example, excluding or including a single seedling would change estimated seedling density by 3183 seedlings ha⁻¹, compared to a change of only 25 seedlings ha⁻¹ in one of Soderberg et al.’s (2024) 11.35-m radius plots.

Neither the lead author of Hanson et al. (2024a) nor the paper’s designated data contact responded to our April 2024 email queries regarding (1) the kind of GPS unit they used, (2) their realized GPS measurement error in the field, and (3) how final plot centers were determined once field crew members were within the GPS unit’s measurement error of their a priori plot center. Even if Hanson et al. used a mapping-grade GPS unit (which we think is unlikely), location error would likely have averaged roughly one half of plot radius. If they instead used a typical hand-held GPS unit, location error would likely have averaged 5 to 10 m or more, or at least 5- to tenfold greater than their plot radius. Thus, unless explicit rules were followed to ensure that final plot placement was unbiased (no such rules were mentioned in Hanson et al. 2024a), field crew members would have had substantial latitude in placing the final plot centers.

Sequoia seedlings often exhibit extreme density variations at scales of <10 m and even <1 m (e.g., Harvey et al. 1980, Demetry 1995, Stephenson et al. 2024; also see Figs. 9 and S2). A consequence of this substantial fine-scale variation is that a small shift in the final location of one of Hanson et al.’s (2024a) 1-m radius plots could mean the difference between recording 0 or >100,000 seedlings ha⁻¹. As a concrete example, a 1-m radius plot we placed on the left side of the scene shown in Fig. 9 yielded 0 seedlings ha⁻¹, whereas a plot we placed on the right side contained 70 seedlings and thus yielded an estimated 222,817 seedlings ha⁻¹. Seemingly innocuous choices for final plot placement—such as avoiding large rocks or tree trunks—could bias results, as could other unconscious or conscious plot-placement choices by field crew members (Elzinga et al. 1998). But potential for plot-placement bias arising from extreme fine-scale variation is greatly reduced in large plots (Elzinga et al. 1998), such as those of Soderberg et al. (2024), each of which covered at least 129-fold more area than one of Hanson et al.’s 1-m radius plots.

Finally, plot-placement bias also could have arisen from Hanson et al.’s (2024a) decision to sample along linear transects, especially if one or more of the transects followed physiographic features that affect sequoia seedling density. For example, two of Hanson et al.’s (2024a) transects appeared to closely follow the bottomlands of the main creek that drains most of Redwood Mountain Grove. Bias could be introduced if bottomlands are prone to different seedling densities than uplands. In contrast, the GRTS sampling design used by Soderberg et al. (2024) is not subject to this form of bias (Stevens and Olsen 2004).

Other potential sources of bias

We considered four additional potential sources of bias (see the Supplementary Information): (1) seasonal timing of censuses, (2) Hanson et al.’s (2024a) use of variable-radius plots, (3) plot edge effects, and (4) trampled seedlings. We deemed none to be sufficient to account for the dramatic difference in the mean seedling densities reported by Soderberg et al. (2024) and Hanson et al. (2024a) (see the Supplementary Information).

Conclusion: potential sources of bias

After considering several potential sources of bias in both data sets, we believe that the sixfold greater seedling densities reported by Hanson et al. (2024a) may have largely resulted from one or more sources of plot-placement bias during Hanson et al.’s sampling. However, because neither the lead author of Hanson et al. (2024a) nor the paper’s designated data contact responded to our queries about their methods, we have been unable to confirm or refute our interpretation.

Hanson et al.’s (mis)use of medians

Hanson et al. (2024a, 2024b) said that they reported median seedling densities but not mean seedling densities because the density data were heavily skewed, and that the use of means rather than medians by Stephenson et al. (2023) meant that important management decisions “are being made in the absence of scientific evidentiary support” (Hanson et al. 2024b).

But as in any study, the appropriate metric depends on the question. For example, if our question was how many seedlings are likely to be found in a given seedling plot of a given size (see below), then medians would be useful. Likewise, if a study’s seedling plot size has been explicitly designed to be at a scale relevant to potential desired planting densities, consideration of medians can be useful to planting decisions (e.g., Welch et al. 2016). But the ultimate question of interest addressed by Stephenson et al. (2023, 2024) and Soderberg et al. (2024) focused on the landscape scale, not the plot scale: what are seedling abundances across broad landscapes? Mean densities are appropriate metrics for the latter question, whereas median densities are not.

One of the reasons medians are inappropriate for characterizing seedling abundances across broad landscapes—namely, because medians vary with plot size—was acknowledged by the first two authors of Hanson et al. (2024a, 2024b) in an earlier publication (Hanson and Chi 2021). We offer a concrete example of the problem, using data from Demetry’s (1995, 1998) study of sequoia seedlings in fire-created forest gaps. Within her largest study gap (Gap 1, spanning 1.164 ha), Demetry used a total station to precisely survey the locations of all 2364 giant sequoia seedlings that were >10 cm tall, 10 years post-fire (Fig. S2). For each of 1000 computer iterations, we sampled seedlings in the gap with 15 randomly located, non-overlapping circular plots and calculated mean and median seedling densities for each iteration, and then averaged results of the 1000 iterations (see the Supplementary Information). The process was repeated for plot radii ranging from 1 m (the size of the majority of Hanson et al.’s (2024a) plots) to 11 m (approximating the 11.35-m radius of the majority of Soderberg et al.’s (2024) plots).

As we expected, the average median seedling density increased continuously with plot size, from ~3 seedlings ha⁻¹ (for the 1-m radius plots) up to ~730 seedlings ha⁻¹ (for the 11-m radius plots) (Fig. 10). The median would continue to increase for plots with radii >11 m until, when the entire gap became the equivalent of a single large plot, the median would finally converge on the study area’s actual seedling density of 2031 ha⁻¹.

Importantly, the increase in median seedling density with plot size is unrelated to cumulative area sampled by the plots. For example, imagine a 1-m grid that samples the entire forest gap illustrated in Fig. S2. A majority of those grid cells would contain no sequoia seedlings (because, at the 1-m scale, most of the gap is empty space), yielding a median density of zero. Now imagine a 20-m grid that samples the entire gap. A majority of those grid cells would contain at least one seedling, thus yielding a non-zero median, even though the same cumulative area was sampled.

Thus, if we were to attempt to use medians to characterize the abundance of regeneration across our study landscapes, all we could say with relative confidence is that, compared to a median based on small plots, a median based on large plots is likely to be closer to representing the actual landscape-scale seedling density, but also probably underestimates that density. In sharp contrast, and regardless of plot size (as previously acknowledged by Hanson and Chi (2021)), the mean density of the plots provides an unbiased estimator of actual seedling density across the study landscape (Fig. 10).

Variation in median seedling density with plot size also underscores the fact that comparisons of medians between data sets based on different plot sizes and (or) variable plot sizes must be interpreted with extreme caution. Hanson et al. (2024b) made such a mismatched comparison of medians between their Nelder Grove seedling densities and those reported by Stephenson et al. (2023).

Finally, Hanson et al. (2024b) inappropriately compared median seedling densities from only the subset of their Nelder Grove plots that burned at high severity with all reference plots in Stephenson et al.’s (2023) data set, which was dominated by areas that burned at low to moderate severity (Stephenson et al. 2024)—effectively an apples-and-oranges comparison of medians. Arguably, a more appropriate comparison would have been with Stephenson et al.’s (2023, 2024) four reference plots that had the highest seedling densities, which likely represent areas that experienced local high-severity fire (Stephenson et al. 2024) (the Stephenson et al. plot data, which extend back to 1970, did not have associated RdNBR data). Stephenson et al.’s four densest fifth-year reference density plots have more than an order of magnitude greater seedling densities (mean = 69,210 seedlings ha⁻¹) than Hanson et al.’s four densest sixth-year Nelder Grove plots (mean = 4902 seedlings ha⁻¹). Even if we instead conservatively compare means of the full data sets, we find that mean seedling density at Nelder Grove 6 years post-fire (1716 ha⁻¹) was fivefold lower than the 5-year post-fire mean reference density (8601 ha⁻¹) reported by Stephenson et al. (2023, 2024), even though the Stephenson et al. reference densities are inherently conservative because they were dominated by low- and moderate-severity fire (Stephenson et al. 2024). Furthermore, if Hanson et al.’s (2024b) Nelder Grove seedling densities were inflated by plot-placement bias (see the preceding subsection), the difference in actual densities would be even larger. Regardless, and contrary to Hanson et al.’s (2024b) apparent contention, available reference densities highlight a potential for inadequate post-fire regeneration at Nelder Grove.

Unsupported and erroneous conclusions

Hanson et al.’s apparent misunderstanding of giant sequoia fire ecology, coupled with methodological problems we have highlighted, appear to have contributed to several of their unsupported and erroneous claims and conclusions. For example, Hanson et al. (2024a) claimed their results “call into question current projects and plans designed to prevent high-severity fire in sequoia groves.” Despite the extensive literature on the topic, the authors apparently did not understand the distinction between the ancient fires that maintained sequoia groves for millennia (i.e., predominantly low to moderate severity surface fires with scattered small patches of higher severity fire that created small forest gaps, in which most regeneration occurred and in which most mature sequoias survived) and the recent wildfires that, in the summers of 2020 and 2021 alone, killed an estimated 13–19% of the world’s ancient sequoias, often in contiguous areas of high sequoia mortality that were >10 ha and even > 100 ha (Shive et al. 2021; Stephenson and Brigham 2021). Simply put, giant sequoia reproduction does not require—and in the past did not require—fires that kill large fractions of ancient sequoias (Stephenson 1996, York et al. 2013; also see the section on “Giant sequoia fire ecology”). In fact, where recent wildfires have burned with historically unprecedented severity (i.e., as crown fires), sequoia reproduction has been well below levels of the past (Soderberg et al. 2024).

Specific to Redwood Mountain Grove, Hurstak (2001) found that before the KNP Complex wildfire >10% of the grove’s area was occupied by canopy gaps of the sizes most conducive to giant sequoia regeneration (i.e., gaps >0.1 ha; Stephenson 1996). Mean gap size in areas that had burned within the last few decades (mostly in prescribed fires) was ~0.3 ha, and those gaps usually contained abundant sequoia regeneration (personal observations; also see Appendix B of Stephenson et al. 2024). That is, Redwood Mountain Grove had decades worth of recent sequoia regeneration—without the deaths of large fractions of its mature sequoias—because prescribed fires had filled the role of ancient fires, triggering successful reproduction in small forest gaps. In contrast, the KNP Complex wildfire completely killed what was likely the largest, densest patch of that regeneration, plus the seed trees that had initially spawned the regeneration (Fig. 7).

Regardless, Hanson et al. (2024a) further stated that Stephenson et al. (2023) and Soderberg et al. (2024) “promoted predominantly low-severity prescribed fire as a more beneficial management approach [than high-severity fire] for giant sequoia groves,” and seemed to imply that this was ill-advised. But their statement is false simply because neither publication promoted, advocated, encouraged, or suggested any approach to management; the papers assessed post-fire seedling densities. Similarly, Hanson et al. (2024b) said that Shive et al. (2022) recommended “lower-severity prescribed fires” but “did not sufficiently address how the generally low levels of sequoia reproduction in a lower-severity fire regime would effectively replenish sequoia grove populations ….” Hanson et al. (2024b) apparently misunderstood the broader context of Shive et al. (2022), where “lower severity” meant lower severity than the recent, historically severe wildfires. Lower severity would thus include, in Shive et al.’s (2022) own words, “low to moderate severity fire, with some smaller patches of high intensity fire that created canopy gaps necessary for regeneration.”

Hanson et al. (2024a, 2024b) further claimed that neither Stephenson et al. (2023) nor Soderberg et al. (2024) investigated whether there is a post-fire sequoia seedling density threshold below which reproduction failure is likely. However, Stephenson et al. (2023) (also see Stephenson et al. 2024) calculated post-fire sequoia seedling reference densities from seedling censuses in 42 plots in eight different sequoia groves—which burned in 26 different fires spanning a 48-year period—and broadly assessed the validity of the reference densities, which they concluded were inherently conservative. Soderberg et al. (2024), in turn, conducted seedling censuses in 116 large, randomly located plots collectively spanning 5.9 ha in four sequoia groves that burned in recent wildfires, showing that seedling densities in the groves were significantly (and sometimes dramatically) lower than the reference densities. Rather than Hanson et al. simply claiming that these approaches did not address the adequacy of post-fire regeneration, the basic standards of scientific discourse would require Hanson et al. to support their claim with evidence.

Hanson et al. additionally stated that their results called into question plans to conduct “widespread planting of sequoia seedlings in high-severity fire areas” (Hanson et al. 2024a) and that they believed that “forest management interventions to facilitate giant sequoia reproduction are unwarranted and potentially counter-productive” (Hanson et al. 2024b). But again, Hanson et al. simply offered no supporting evidence—whether from demographic modeling or quantitative comparisons with seedling reference densities—that planting might be unwarranted, nor did they offer any evidence refuting the contradictory findings of Soderberg et al. (2024) (see the preceding paragraph).

Hanson et al. (2024a, 2024b) presented several photographs of high densities of sequoia seedlings within high-severity burn areas of both Redwood Mountain Grove and Nelder Grove, perhaps to bolster their suggestion that natural regeneration was sufficient and seedling planting was thus unwarranted. However, our conversations with sequoia managers have indicated that areas with dense regeneration —often where the mature sequoias were killed by crown scorch but not crown fire (Fig. 1)—are not targets for replanting (C. M. Brigham, National Park Service, Three Rivers, California, USA, oral communication, 30 July 2024). Instead, replanting focuses on the large areas with little or no reproduction, often found where crown fires occurred (Fig. 11). Hanson et al. (2024a, 2024b) simply did not acknowledge the role of crown fire and historically exceptional drought in reducing post-fire reproduction, nor the existence of large areas having little post-fire reproduction.

To support their expressed doubts and concerns about the efficacy and consequences of mechanical fuel treatments, Hanson et al. (2024b) cited papers that suggest (to us, at least) that Hanson et al. conflated mechanical fuel reductions (which may cut some small trees while leaving large trees) with logging (removal of large trees). Regardless, a recent meta-analysis (40 studies spanning 172 sites) of fuel treatments in dry forests of the western United States (including California mixed conifer forests; also see Brodie et al. 2024) unequivocally demonstrated that, relative to untreated controls, mechanical fuel treatments followed by pile burning or prescribed fire substantially reduced the severity of subsequent wildfires (Davis et al. 2024).

Conclusions

Hanson et al.’s (2024a, 2024b) apparent unfamiliarity with (and misinterpretation of) the broad literature on giant sequoia fire ecology, coupled with their methodological problems, may have contributed to their flawed and inaccurate framing of sequoia conservation issues and their unsupported and erroneous claims and conclusions. Hanson et al. (2024b) called the lack of sequoia regeneration associated with long-term fire exclusion a “significant conservation threat,” yet failed to acknowledge the acute conservation threat posed by the recent wildfire deaths of 13–19% of seed-bearing sequoias across their native range. That is, for the first time in at least a millennium the continued existence of giant sequoias in several areas now depends largely upon the survival of a cohort of small seedlings—whether naturally seeded or planted—in the face of continued climatic changes and increasingly severe wildfires. Figuratively speaking, within those areas all the eggs are now in one vulnerable basket (cf. Figure 7).

Hanson et al. (2024a) further stated that their observations of natural post-fire sequoia regeneration in Redwood Mountain Grove were “encouraging for the conservation of giant sequoias.” However, three things undermine this claim. First, the sixfold higher seedling densities found by Hanson et al. (2024a) compared to Soderberg et al. (2024) in the grove may have largely resulted from one or more forms of plot-placement bias (see “Assessing apparent data bias,” above), suggesting that Hanson et al. may have substantially overestimated actual seedling densities across the landscape. Second, Hanson et al. (2024a) failed to offer any evidence that their observed seedling densities would be sufficient to regenerate the forest. Finally, Soderberg et al.’s (2024) order-of-magnitude more robust sampling of the same grove found post-fire seedling densities significantly lower than historical post-fire densities, suggesting inadequate natural regeneration.

The sizable existing literature, plus analyses we have presented here, show that giant sequoias successfully regenerated for millennia under a surface fire regime in which the vast majority of mature sequoias survived. That is, sustaining sequoia populations does not—and never did—depend on the deaths of most or all ancient sequoias over broad areas, as has occurred in recent wildfires. Key issues confronting sequoia stewards today thus include extensive losses of seed-bearing sequoias often followed by inadequate natural regeneration (Shive et al. 2022; Soderberg et al. 2024), coupled with the potential for reduced seedling survival in the face of continued increases in temperature, severe drought, and extreme wildfire (Stephenson et al. 2024).

Availability of data and materials

Data used to determine the optimal trunk diameter for distinguishing millennial and non-millennial sequoias are available at https://doi.org/10.5066/P14GGCVN. Data used to produce Table 1 and the maps in Figs. 4 and S1 can be obtained upon request to author J. A. Flickinger (joshua_flickinger@nps.gov).

References

Andrews, E. D. 2013. A natural resource condition assessment for Sequoia and Kings Canyon National Parks: Appendix 7b – hydrology of Sierra Nevada Network parks. Natural Resource Report NPS/SEKI/NRR—2013/665.7b. Fort Collins, Colorado: National Park Service.
Google Scholar
Baker, W. L. 2014. Historical forest structure and fire in Sierran mixed‐conifer forests reconstructed from General Land Office survey data. Ecosphere 5 (7): 1–70.
Baker, W. L., and C. T. Hanson. 2017. Improving the use of early timber inventories in reconstructing historical dry forests and fire in the western United States. Ecosphere 8 (9): e01935.
Bellows, R. S., A. C. Thomson, K. J. Helmstedt, R. A. York, and M. D. Potts. 2016. Damage and mortality patterns in young mixed conifer plantations following prescribed fires in the Sierra Nevada, California. Forest Ecology and Management 376: 193–204.
Article Google Scholar
Bonnicksen, T. M., and E. C. Stone. 1982. Reconstruction of a presettlement giant sequoia-mixed conifer forest community using the aggregation approach. Ecology 63 (4): 1134–1148.
Article Google Scholar
Brodie, E. G., E. E. Knapp, W. R. Brooks, S. A. Drury, and M. W. Ritchie. 2024. Forest thinning and prescribed burning treatments reduce wildfire severity and buffer the impacts of severe fire weather. Fire Ecology 20 (1): 17.
Article Google Scholar
Brown, P. T., H. Hanley, A. Mahesh, C. Reed, S. J. Strenfel, S. J. Davis, A. K. Kochanski, and C. B. Clements. 2023. Climate warming increases extreme daily wildfire growth risk in California. Nature 621 (7980): 760–766.
Article CAS PubMed Google Scholar
Buchholz, J. T. 1938. Cone formation in Sequoia gigantea. I. The relation of stem size and tissue development to cone formation. II. The history of the seed cone. American Journal of Botany 25 (4): 296–305.
Article Google Scholar
Caprio, A. C., L. S. Mutch, T. W. Swetnam, and C. H. Baisan. 1994. Temporal and spatial patterns of giant sequoia radial growth response to a high severity fire in A.D. 1297. Report to the California Department of Forestry and Fire Protection, Mountain Home State Forest, Contract No. 8CA17025.
Clark, J. S., M. Silman, R. Kern, E. Macklin, and J. HilleRisLambers. 1999. Seed dispersal near and far: Patterns across temperature and tropical forests. Ecology 80 (5): 1475–1494.
Article Google Scholar
Collins, B. M. 2014. Fire weather and large fire potential in the northern Sierra Nevada. Agricultural and Forest Meteorology 189: 30–35.
Article Google Scholar
Collins, B. M., and G. B. Roller. 2013. Early forest dynamics in stand-replacing fire patches in the northern Sierra Nevada, California, USA. Landscape Ecology 28 (9): 1801–1813.
Article Google Scholar
Coop, J. D., S. A. Parks, C. S. Stevens-Rumann, S. D. Crausbay, P. E. Higuera, M. D. Hurteau, A. Tepley, E. Whitman, T. Assal, B. M. Collins, K. T. Davis, S. Dobrowski, D. A. Falk, P. J. Fornwalt, P. Z. Fulé, B. J. Harvey, V. R. Kane, C. E. Littlefield, E. Q. Margolis, M. North, M. -A. Parisien, S. Prichard, and K. C. Rodman. 2020. Wildfire-driven forest conversion in western North landscapes. BioScience 70 (8): 659–673.
Article PubMed PubMed Central Google Scholar
Coppoletta, M., K.E. Merriam, and B.M. Collins. 2016. Post-fire vegetation and fuel development influences fire severity patterns in reburns. Ecological Applications 26 (3): 686–699.
Article PubMed Google Scholar
Davis, K. T., M. D. Robles, K. B. Kemp, P. E. Higuera, T. Chapman, K. L. Metlen, J. L. Peeler, K. C. Rodman, T. Woolley, R. N. Addington, B. J. Buma, C. A. Cansler, M. J. Case, B. M. Collins, J. D. Coop, S. Z. Dobrowski, N. S. Gill, C. Haffey, L. B. Harris, B. J. Harvey, R. D. Haugo, M. D. Hurteau, D. Kulakowski, C. E. Littlefield, L. A. McCauley, N. Povak, K. L. Shive, E. Smith, J. T. Stevens, C. S. Stevens-Rumann, A. H. Taylor, A. J. Tepley, D. J. N. Young, R. A. Andrus, M. A. Battaglia, J. K. Berkey, S. U. Busby, A. R. Carlson, M. E. Chambers, E. K. Dodson, D. C. Donato, W. M. Downing, P. J. Fornwalt, J. S. Halofsky, A. Hoffman, A. Holz, J. M. Iniguez, M. A. Krawchuk, M. R. Kreider, A. J. Larson, G. W. Meigs, J. P. Roccaforte, M. T. Rother, H. Safford, M. Schaedel, J. S. Sibold, M. P. Singleton, M. G. Turner, A. K. Urza, K. D. Clark-Wolf, L. Yocom, J. B. Fontaine, and J. L. Campbell. 2023. Reduced fire severity offers near-term buffer to climate-driven declines in conifer resilience across the western United States. Proceedings of the National Academy of Sciences 120 (11): e2208120120.
Article CAS Google Scholar
Davis, K. T., J. Peeler, J. Fargione, R. D. Haugo, K. L. Metlen, M. D. Robles, and T. Woolley. 2024. Tamm Review: A meta-analysis of thinning, prescribed fire, and wildfire effects on subsequent wildfire severity in conifer dominated forests of the Western US. Forest Ecology and Management 561: 121885.
Article Google Scholar
Demetry, A. 1995. Regeneration patterns within canopy gaps in a giant sequoia - mixed conifer forest: implications for forest restoration. M.S. thesis, Northern Arizona University, Flagstaff.
Demetry, A. 1998. A natural disturbance model for the restoration of Giant Forest Village, Sequoia National Park. In Proceedings of High Altitude Revegetation Workshop, No. 13 (W. R. Keammerer and E. F. Redente, editors). Fort Collins: Colorado Water Resources Research Institute, Information Series No. 89.
Elzinga, C. L., D. W. Salzer, and J. W. Willoughby. 1998. Measuring and Monitoring Plant Populations. BLM Technical Reference 1730–1, U.S. Government Printing Office. 477 pages.
Fites-Kaufman, J. A., P. Rundel, N. Stephenson, and D. A. Weixelman. 2007. Montane and subalpine vegetation of the Sierra Nevada and Cascade ranges. Pages 456–501 in M. Barbour, T. Keeler-Wolf, and A. A. Schoenherr (eds.), Terrestrial Vegetation of California, 3rd edition. University of California Press.
Gill, N. S., M. G. Turner, C. D. Brown, S. I. Glassman, S. L. Haire, W. D. Hansen, E. R. Pansing, S. B. St, and Clair, and D. F. Tomback. 2022. Limitations to propagule dispersal will constrain postfire recovery of plants and fungi in western coniferous forests. BioScience 72 (4): 347–364.
Article Google Scholar
Gleason, K. E., J. R. McConnell, M. M. Arienzo, N. Chellman, and W. M. Calvin. 2019. Four-fold increase in solar forcing on snow in western US burned forests since 1999. Nature Communications 10 (1): 2026.
Article PubMed PubMed Central Google Scholar
Gonzalez, P. 2012. Climate change trends and vulnerability to biome shifts in the southern Sierra Nevada. U. S. National Park Service, Climate Change Response Program, Washington, D. C. 37 pages.
Goss, M., D. L. Swain, J. T. Abatzoglou, A. Sarhadi, C. A. Kolden, A. P. Williams, and N. S. Diffenbaugh. 2020. Climate change is increasing the likelihood of extreme autumn wildfire conditions across California. Environmental Research Letters 15 (9): 094016.
Article Google Scholar
Greene, D. F., J. M. Kane, E. Pounden, and S. T. Michaletz. 2024. Cone allometry and seed protection from fire are similar in serotinous and nonserotinous conifers. New Phytologist. 242 (1): 93–106.
Article CAS PubMed Google Scholar
Guiterman, C. H., R. M. Gregg, L. A. Marshall, J. J. Beckmann, P. J. van Mantgem, D. A. Falk, J. E. Keeley, A. C. Caprio, J. D. Coop, P. J. Fornwalt, C. Haffey, R. K. Hagmann, S. T. Jackson, A. M. Lynch, E. Q. Margolis, C. Marks, M. D. Meyer, H. Safford, A. D. Syphard, A. Taylor, C. Wilcox, D. Carril, C. A. F. Enquist, D. Huffman, J. Iniguez, N. A. Molinari, C. Restaino, and J. T. Stevens. 2022. Vegetation type conversion in the US Southwest: Frontline observations and management responses. Fire Ecology 18 (1): 1–16.
Article Google Scholar
Gutierrez, A. A., S. Hantson, B. Langenbrunner, B. Chen, Y. Jin, M. L. Goulden, and J. T. Randerson. 2021. Wildfire response to changing daily temperature extremes in California’s Sierra Nevada. Science Advances 7 (47): eabe6417.
Article PubMed PubMed Central Google Scholar
Hagmann, R. K., J. T. Stevens, J. M. Lydersen, B. M. Collins, J. J. Battles, P. F. Hessburg, C. R. Levine, A. G. Merschel, S. L. Stephens, A. H. Taylor, J. F. Franklin, D. L. Johnson, and K. N. Johnson. 2018. Improving the use of early timber inventories in reconstructing historical dry forests and fire in the western United States: Comment. Ecosphere 9 (7): e02232. https://doi.org/10.1002/ecs2.2232.
Article Google Scholar
Hagmann, R. K., P. F. Hessburg, S. J. Prichard, N. A. Povak, P. M. Brown, P. Z. Fulé, R. E. Keane, E. E. Knapp, J. M. Lydersen, K. L. Metlen, M. J. Reilly, A. J. Sánchez Meador, S. L. Stephens, J. T. Stevens, A. H. Taylor, L. L. Yocom, M. A. Battaglia, D. J. Churchill, L. D. Daniels, D. A. Falk, P. Henson, J. D. Johnston, M. A. Krawchuk, C. R. Levine, G. W. Meigs, A. G. Merschel, M. P. North, H. D. Safford, T. W. Swetnam, and A. E. M. Waltz. 2021. Evidence for widespread changes in the structure, composition, and fire regimes of western North American forests. Ecological Applications 31 (8): e02431.
Article CAS PubMed PubMed Central Google Scholar
Halpin, P. N. 1995. A Cross-scale Analysis of Environmental Gradients and Forest Pattern in the Giant Sequoia – Mixed Conifer Forest of the Sierra Nevada. Dissertation, University of Virginia. 277 pages.
Hanson, C. T., and T. Y. Chi. 2021. Impacts of postfire management are unjustified in spotted owl habitat. Frontiers in Ecology and Evolution 9: 596282.
Article Google Scholar
Hanson, C. T., T. Y. Chi, M. Khosla, B. C. Baker, and C. Swolgaard. 2024a. Reproduction of a serotinous conifer, the giant sequoia, in a large high-severity fire area. Fire 7 (2): 44.
Article Google Scholar
Hanson, C. T., T. Y. Chi, B. C. Baker, M. Khosla, and M. K. Dorsey. 2024b. Postfire reproduction of a serotinous conifer, the giant sequoia, in the Nelder Grove. California. Ecology and Evolution 14 (4): e11213.
Article PubMed PubMed Central Google Scholar
Harpold, A. A., and N. P. Molotch. 2015. Sensitivity of soil water availability to changing snowmelt timing in the western U.S. Geophysical Research Letters 42 (19): 8011–8020.
Article Google Scholar
Harris, L., and A. H. Taylor. 2017. Previous burns and topography limit and reinforce fire severity in a large wildfire. Ecosphere 8 (11): e02019.
Article Google Scholar
Hartesveldt, R. J. 1964. Fire ecology of the giant sequoias. Natural History 73 (10): 12–19.
Google Scholar
Hartesveldt, R. J., H. T. Harvey, H. S. Shellhammer, and R. E. Stecker. 1975. The Giant Sequoia of the Sierra Nevada. Washington, DC: USDI National Park Service.
Google Scholar
Harvey, H. T., H. S. Shellhammer, and R. E. Stecker. 1980. Giant Sequoia Ecology. Washington, DC: USDI National Park Service.
Google Scholar
Hatchett, B. J., A. L. Koshkin, K. Guirguis, K. Rittger, A. W. Nolin, A. Heggli, A. M. Rhoades, A. East, E. R. Siirila-Woodburn, W. T. Brandt, A. Gershunov, and K. Haleakala. 2023. Midwinter dry spells amplify post-fire snowpack decline. Geophysical Research Letters, e2022GL101235.
Hurstak, R. 2001. A comparative analysis of canopy gap distributions in two giant sequoia groves. Thesis: California Polytechnic State University, San Luis Obispo.
Google Scholar
Huntington, E. 1914. The Climatic Factor as Illustrated in Arid America. Carnegie Institute of Washington, Publ. 192. 341 pages.
Keeley, J. E., and P. Zedler. 1998. Evolution of life histories in Pinus. Chapter 12 (pages 219–249) in Richardson, D. M. (editor), Ecology and Biogeography of Pinus. Cambridge University Press.
Keifer, M. 1998. Fuel load and tree density changes following prescribed fire in the giant sequoia – mixed conifer forest: The first 14 years of fire effects monitoring. Proceedings of the Tall Timbers Fire Ecology Conference 20: 306–309.
Google Scholar
Keifer, M., J. W. van Wagtendonk, and M. Buhler. 2006. Long-term surface fuel accumulation in burned and unburned mixed-conifer forests of the central and southern Sierra Nevada, CA (USA). Fire Ecology 2: 53–72.
Article Google Scholar
Kilgore, B. M., and R. W. Sando. 1975. Crown-fire potential in a sequoia forest after prescribed burning. Forest Science 21 (1): 83–87.
Google Scholar
Kilgore, B. M., and D. Taylor. 1979. Fire history of a sequoia-mixed conifer forest. Ecology 60 (1): 129–142.
Article Google Scholar
Lamont, B. B. 1991. Canopy seed storage and release: What’s in a name? Oikos 60 (2): 266–268.
Article Google Scholar
Lamont, B. B., D. C. Le Maitre, R. M. Cowling, and N. J. Enright. 1991. Canopy seed storage in woody plants. Botanical Review 57 (4): 277–317.
Article Google Scholar
Levine, C. R., C. V. Cogbill, B. M. Collins, A. J. Larson, J. A. Lutz, M. P. North, C. M. Restaino, H. D. Safford, S. L. Stephens, and J. J. Battles. 2017. Evaluating a new method for reconstructing forest conditions from General Land Office survey records. Ecological Applications 27 (5): 1498–1513.
Article PubMed Google Scholar
Levine, C. R., C. V. Cogbill, B. M. Collins, A. J. Larson, J. A. Lutz, M. P. North, C. M. Restaino, H. D. Safford, S. L. Stephens, and J. J. Battles. 2019. Estimating historical forest density from land-survey data: A response to Baker and Williams (2018). Ecological Applications. https://doi.org/10.1002/eap.1968.
Lydersen, J. M., B. M. Collins, M. L. Brooks, J. R. Matchett, K. L. Shive, N. A. Povak, V. R. Kane, and D. F. Smith. 2017. Evidence of fuels management and fire weather influencing fire severity in an extreme fire event. Ecological Applications 27 (7): 2013–2030.
Article PubMed Google Scholar
Lydersen, J. M., B. M. Collins, M. Coppoletta, M. R. Jaffe, H. Northrop, and S. L. Stephens. 2019. Fuel dynamics and reburn severity following high-severity fire in a Sierra Nevada, USA, mixed-conifer forest. Fire Ecology 15 (1): 43.
Article Google Scholar
Martín-Sanz, R. C., M. Callejas-Díaz, J. Tonnabel, and J. M. Climent. 2017. Maintenance costs of serotiny in a variably serotinous pine: The role of water supply. PLoS ONE 12 (7): e0181648.
Article PubMed PubMed Central Google Scholar
Meyer, M. D., and H. D. Safford. 2011. Giant sequoia regeneration in groves exposed to wildfire and retention harvest. Fire Ecology 7: 2–16.
Article Google Scholar
Meyer, M. D., T. Sowards, R. Wayman, and B. Bulaon. 2024. Windy Fire Post-fire Ecological Assessment: 2022 Field Inventory Report. USDA Forest Service Region 5 Ecology Program in partnership with USDA Forest Service Forest Health Protection and the University of California – Davis. 88 pages.
Miller, J. D., and A. E. Thode. 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.
Article Google Scholar
Moran, E. V., A. J. Das, J. E. Keeley, and N. L. Stephenson. 2019. Negative impacts of summer heat on Sierra Nevada tree seedlings. Ecosphere 10 (6): e02776.
Article Google Scholar
Mote, P. W., S. Li, D. P. Lettenmaier, M. Xiao, and R. Engel. 2018. Dramatic declines in snowpack in the western US. NPJ Climate and Atmospheric Science 1 (1): 2. https://doi.org/10.1038/s41612-018-0012-1.
Article Google Scholar
Mutch, L. S., and T. W. Swetnam. 1995. Effects of fire severity and climate on ring-width growth of giant sequoia after burning. Pages 241–246 in J. K. Brown, R. W. Mutch, C. W. Spoon, and R. H. Wakimoto (tech. coords.). Proceedings: Symposium on fire in wilderness and park management. USDA Forest Service Gen. Tech. Rep. INT-GTR-320.
Nichols, H. T. 1989. Managing fire in Sequoia and Kings Canyon national parks. Fremontia 16 (4): 11–14.
Google Scholar
North, M. P., J. T. Stevens, D. F. Greene, M. Coppoletta, E. E. Knapp, A. M. Latimer, C. M. Restaino, R. E. Tompkins, K. R. Welch, R. A. York, D. J. N. Young, J. N. Axelson, T. N. Buckley, B. L. Estes, R. N. Hager, J. W. Long, M. D. Meyer, S. M. Ostoja, H. D. Safford, K. L. Shive, C. L. Tubbesing, H. Vice, D. Walsh, C. M. Werner, and P. Wyrsch. 2019. Tamm Review: Reforestation for resilience in dry western US forests. Forest Ecology and Management 432: 209–224.
Article Google Scholar
Prichard, S. J., P. F. Hessburg, R. K. Hagmann, N. A. Povak, S. Z. Dobrowski, M. D. Hurteau, and M. D., V. R. Kane, R. E. Keane, L. N. Kobziar, C. A. Holden, M. North, S. A. Parks, H. D. Safford, J. T. Stevens, L. L. Yocom, D. J. Churchill, R. W. Gray, D. W. Huffman, F. K. Lake, and P. Khatri-Chhetri. 2021. Adapting western North American forests to climate change and wildfires: 10 common questions. Ecological Applications 31 (8): e02433.
Article PubMed PubMed Central Google Scholar
Reilly, M. J., V. J. Monleon, E. S. Jules, and R. J. Butz. 2019. Range-wide population structure and dynamics of a serotinous conifer, knobcone pine (Pinus attenuata L.), under an anthropogenically-altered disturbance regime. Forest Ecology and Management 441: 182–191.
Article Google Scholar
Rother, D. E., F. De Sales, D. Stow, and J. P. McFadden. 2022. Summer and fall extreme fire weather projected to occur more often and affect a growing portion of California throughout the 21st century. Fire 5 (6): 177.
Article Google Scholar
Rundel, P. W. 1972. Habitat restriction in giant sequoia: The environmental control of grove boundaries. American Midland Naturalist 87 (1): 81–99.
Article Google Scholar
Schubert, G. H., and N. M. Beetham. 1962. Silvical characteristics of giant sequoia. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Technical Paper 20. 16 pages.
Schwilk, D. W., and D. D. Ackerly. 2001. Flammability and serotiny as strategies: Correlated evolution in pines. Oikos 94 (2): 326–336.
Article Google Scholar
Shive, K. L., H. K. Preisler, K. R. Welch, H. D. Safford, R. J. Butz, K. L. O’Hara, and S. L. Stephens. 2018. From the stand scale to the landscape scale: Predicting the spatial patterns of forest regeneration after disturbance. Ecological Applications 28: 1626–1639.
Article PubMed Google Scholar
Shive, K., C. Brigham, T. Caprio, and P. Hardwick. 2021. 2021 Fire Season Impacts to Giant Sequoias. https://www.nps.gov/articles/000/2021-fire-season-impacts-to-giant-sequoias.htm. Accessed 11 Jan 2024.
Shive, K. L., A. Wuenschel, L. J. Hardlund, S. Morris, M. D. Meyer, and S. M. Hood. 2022. Ancient trees and modern wildfires: Declining resilience to wildfire in the highly fire-adapted giant sequoia. Forest Ecology and Management 511: 120110.
Article Google Scholar
Sillett, S. C., R. Van Pelt, A. L. Carroll, R. D. Kramer, A. R. Ambrose, and D. A. Trask. 2015. How do tree structure and old age affect growth potential of California redwoods? Ecological Monographs 85 (2): 181–212.
Article Google Scholar
Smoot, E. E., and K. E. Gleason. 2021. Forest fires reduce snow-water storage and advance the timing of snowmelt across the Western U.S. Water 13 (24): 3533.
Article Google Scholar
Soderberg, D. N., A. J. Das, N. L. Stephenson, M. D. Meyer, C. A. Brigham, and J. Flickinger. 2024. Assessing giant sequoia mortality and regeneration following high severity wildfire. Ecosphere 15 (3): e4789.
Article Google Scholar
Stark, N. 1968. The environmental tolerance of the seedling stage of Sequoiadendron giganteum. American Midland Naturalist 80: 84–95.
Article Google Scholar
Stephens, S. L., and M. A. Finney. 2002. Prescribed fire mortality of Sierra Nevada mixed conifer tree species: Effects of crown damage and forest floor combustion. Forest Ecology and Management 162 (2–3): 261–271.
Article Google Scholar
Stephenson, N. L. 1994. Long-term dynamics of giant sequoia populations: implications for managing a pioneer species. Pages 56–63 in P. S. Aune (tech. coord.), Proceedings of the Symposium on Giant Sequoias: their place in the ecosystem and society. USDA Forest Service Gen. Tech. Rep. PSW-151.
Stephenson, N. L. 1996. Ecology and management of giant sequoia groves. Pages 1431–1467 in Sierra Nevada Ecosystem Project: Final Report to Congress, vol. II, Assessments and scientific basis for management options. Centers for Water and Wildland Resources, University of California, Davis.
Stephenson, N. L. 1999. Reference conditions for giant sequoia forest restoration: Structure, process, and precision. Ecological Applications 9: 1253–1265.
Article Google Scholar
Stephenson, N., and C. Brigham. 2021. Preliminary estimates of sequoia mortality in the 2020 Castle Fire. https://www.nps.gov/articles/000/preliminary-estimates-of-sequoia-mortality-in-the-2020-castle-fire.htm. Accessed 11 January 2024.
Stephenson, N. L., D. J. Parsons, and T. W. Swetnam. 1991. Restoring natural fire to the sequoia-mixed conifer forest: Should intense fire play a role? Proceedings of the Tall Timbers Fire Ecology Conference 17: 321–337.
Google Scholar
Stephenson, N. L., A. C. Caprio, D. N. Soderberg, A. J. Das, E. L. Lopez, and A. P. Williams. 2023. Post-fire reference densities for giant sequoia seedlings. EcoEvoRxiv: https://ecoevorxiv.org/repository/view/5457/ (https://doi.org/10.32942/X2X88S).
Stephenson, N. L., A. C. Caprio, D. N. Soderberg, A. J. Das, E. L. Lopez, and A. P. Williams. 2024. Post-fire reference densities for giant sequoia seedlings in a new era of high-severity wildfires. Forest Ecology and Management 562: 121916.
Article Google Scholar
Stevens, J. T. 2017. Scale-dependent effects of post-fire canopy cover on snowpack depth in montane coniferous forests. Ecological Applications 27: 1888–1900.
Article PubMed Google Scholar
Stevens, J. T., M. M. Kling, D. W. Schwilk, J. M. Varner, and J. M. Kane. 2020. Biogeography of fire regimes in western U.S. conifer forests: a trait-based approach. Global Ecology and Biogeography 29 (5): 944–955.
Article Google Scholar
Stevens, D. L., Jr., and A. R. Olsen. 2004. Spatially balanced sampling of natural resources. Journal of the American Statistical Association 99 (465): 262–278.
Article Google Scholar
Stevens-Rumann, C. S., and P. Morgan. 2019. Tree regeneration following wildfires in the western US: A review. Fire Ecology 15 (1): 1–17.
Article Google Scholar
Stevens-Rumann, C. S., S. J. Prichard, E. Whitman, M. A. Parisien, and A. J. Meddens. 2022. Considering regeneration failure in the context of changing climate and disturbance regimes in western North America. Canadian Journal of Forest Research 52: 1281–1302.
Article Google Scholar
Stewart, J. A. E., P. J. van Mantgem, D. J. N. Young, K. L. Shive, H. K. Preisler, A. J. Das, N. L. Stephenson, J. E. Keeley, H. D. Safford, M. C. Wright, K. R. Welch, and J. H. Thorne. 2021. Influence of postfire climate and seed production on conifer regeneration. Ecological Applications 31 (3): e02280. https://doi.org/10.1002/eap.2280.
Article PubMed Google Scholar
Swetnam, T. W. 1993. Fire history and climate change in giant sequoia groves. Science 262 (5135): 885–889.
Article CAS PubMed Google Scholar
Swetnam, T. W., C. H. Baisan, A. C. Caprio, P. M. Brown, R. Touchan, R. S. Anderson, and D. J. Hallett. 2009. Multi-millennial fire history of the Giant Forest, Sequoia National Park, California, USA. Fire Ecology 5: 120–150.
Article Google Scholar
Taylor, A. H., L. B. Harris, and S. A. Drury. 2021. Drivers of fire severity shift as landscapes transition to an active fire regime, Klamath Mountains, USA. Ecosphere 12 (9): e03734.
Article Google Scholar
Taylor, A. H., L. B. Harris, and C. N. Skinner. 2022. Severity patterns of the 2021 Dixie Fire exemplify the need to increase low-severity fire treatments in California’s forests. Environmental Research Letters 17 (7): 071002.
Article Google Scholar
Tercek, M. T., D. Thoma, J. E. Gross, K. Sherrill, S. Kagone, and G. Senay. 2021. Historical changes in plant water use and need in the continental United States. PLoS ONE 16 (9): e0256586.
Article CAS PubMed PubMed Central Google Scholar
van Mantgem, P. J., L. B. Lalemand, M. Keifer, and J. M. Kane. 2016. Duration of fuels reduction following prescribed fire in coniferous forests of U.S. national parks in California and the Colorado Plateau. Forest Ecology and Management 379: 265–272.
Article Google Scholar
Warren, R., and A. J. Fordham. 1978. The fire pines. Arnoldia 38 (1): 1–11.
Welch, K. R., H. D. Safford, and T. P. Young. 2016. Predicting conifer establishment post wildfire in mixed conifer forests of the North American Mediterranean-climate zone. Ecosphere 7 (12): e01609. https://doi.org/10.1002/ecs2.1609.
Article Google Scholar
Williams, A. P., J. T. Abatzoglou, A. Gershunov, J. Guzman-Morales, D. A. Bishop, J. K. Balch, and D. P. Lettenmaier. 2019. Observed impacts of anthropogenic climate change on wildfire in California. Earth’s Future 7 (8): 892–910. https://doi.org/10.1029/2019EF001210.
Article Google Scholar
Williams, A. P., B. I. Cook, and J. E. Smerdon. 2022. Rapid intensification of the emerging southwestern North American megadrought in 2020–2021. Nature Climate Change 12: 232–234.
Article Google Scholar
York, R. A., N. L. Stephenson, M. Meyer, S. Hanna, T. Moody, T. Caprio, and J. J. Battles. 2013. A natural resource condition assessment for Sequoia and Kings Canyon National Parks: Appendix 11 - Giant sequoia. Natural Resource Report NPS/SEKI/NRR—2013/665.11. National Park Service, Fort Collins, CO. (106 pages).
York, R. A., H. Noble, L. N. Quinn-Davidson, and J. J. Battles. 2021. Pyrosilviculture: Combining prescribed fire with gap-based silviculture in mixed-conifer forests of the Sierra Nevada. Canadian Journal of Forest Research 51 (6): 781–791.
Article CAS Google Scholar

Download references

Acknowledgements

We thank Phil van Mantgem and three anonymous reviewers for helpful comments on the manuscript, and Athena Demetry for supplying her thesis data. This work was funded by the U.S. Geological Survey’s Ecosystems and Climate and Land Use Research and Development programs and by the National Park Service. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Funding

This work was funded by the U.S. Geological Survey’s Ecosystems and Climate and Land Use Research and Development programs and by the National Park Service.

Author information

Authors and Affiliations

U.S. Geological Survey, Western Ecological Research Center, Sequoia-Kings Canyon Field Station, 47050 Generals Highway, Three Rivers, CA, 93271, USA
Nathan L. Stephenson, David N. Soderberg & Adrian J. Das
Division of Resources Management and Science, Sequoia and Kings Canyon National Parks, 47050 Generals Highway, Three Rivers, CA, 93271, USA
Joshua A. Flickinger & Anthony C. Caprio

Authors

Nathan L. Stephenson
View author publications
You can also search for this author in PubMed Google Scholar
David N. Soderberg
View author publications
You can also search for this author in PubMed Google Scholar
Joshua A. Flickinger
View author publications
You can also search for this author in PubMed Google Scholar
Anthony C. Caprio
View author publications
You can also search for this author in PubMed Google Scholar
Adrian J. Das
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

NLS, DNS, ACC, and AJD conceptualized the paper. NLS, DNS, and JAF analyzed data. NLS wrote most of the manuscript and all authors contributed to revisions.

Corresponding author

Correspondence to Nathan L. Stephenson.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

42408_2024_316_MOESM1_ESM.pdf

Additional File 1: Supplementary Text. Supporting text (including references) for the sections on Giant sequoia fire ecology and Methodological problems. Table S1. Data used to determine the optimal trunk diameter for distinguishing giant sequoias ≥1000 years old from those <1000 years old. Table S2. Contingency table of numbers of giant sequoias in Redwood Mountain Grove, California, USA by estimated age and trunk diameter at breast height. Figure S1. Distribution of millennial sequoias (those estimated to be ≥1000 years old) in the 88.5-ha Castle Creek Grove, Sequoia National Park, California, USA. Figure S2. Map of 10-year post-fire sequoia seedlings >10 cm tall (red circles) within Demetry’s (1995, 1998) fire-created Gap 1.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Stephenson, N.L., Soderberg, D.N., Flickinger, J.A. et al. Effects of recent wildfires on giant sequoia groves were anomalous at millennial timescales: a response to Hanson et al.. fire ecol 20, 89 (2024). https://doi.org/10.1186/s42408-024-00316-5

Download citation

Received: 07 July 2024
Accepted: 13 August 2024
Published: 30 September 2024
DOI: https://doi.org/10.1186/s42408-024-00316-5

Effects of recent wildfires on giant sequoia groves were anomalous at millennial timescales: a response to Hanson et al.

Abstract

Background

Results

Conclusions

Resumen

Antecedentes

Resultados

Conclusiones

Introduction

Giant sequoia fire ecology

The nature of past fire regimes and effects

Sequoia reproduction first increases, then declines with increasing fire severity

Crown fires have been absent for at least a millennium

Recent high-severity wildfires were anomalous at millennial timescales

Giant sequoia is maladapted to crown fire

Distance to surviving seed trees reflects vulnerability to future loss

Novel environmental conditions likely reduce seedling survival

Methodological problems

Assessing apparent data bias

Density differences between data sets

Seedling cohorts compared

Plot sizes

Plot-placement bias

Other potential sources of bias

Conclusion: potential sources of bias

Hanson et al.’s (mis)use of medians

Unsupported and erroneous conclusions

Conclusions

Availability of data and materials

References

Acknowledgements

Funding

Author information

Authors and Affiliations

Contributions

Corresponding author

Ethics declarations

Ethics approval and consent to participate

Consent for publication

Competing interests

Additional information

Publisher’s Note

Supplementary Information

42408_2024_316_MOESM1_ESM.pdf

Rights and permissions

About this article

Cite this article

Share this article

Keywords