The ecology of mixed severity fire regimes in Washington, Oregon, and Northern California

Forests characterized by mixed-severity fires occupy a broad moisture gradient between lower elevation forests typified by low-severity fires and higher elevation forests in which high-severity, stand replacing fires are the norm. Mixed-severity forest types are poorly documented and little understood but likely occupy significant areas in the western United States. By definition, mixed-severity types have high beta diversity at meso-scales, encompassing patches of both high and low severity and gradients in between. Studies of mixed-severity types reveal complex landscapes in which patch sizes follow a power law distribution with many small and few large patches. Forest types characterized by mixed severity can be classified according to the modal proportion of high to low severity patches, which increases from relatively dry to relatively mesic site conditions. Mixed-severity regimes are produced by interactions between top-down forcing by climate and bottom-up shaping by topography and the flammability of vegetation, although specific effects may vary widely across the region, especially the relation between aspect and fire severity. History is important in shaping fire behavior in mixed-severity landscapes, as patterns laid down by previous fires can play a significant role in shaping future fires. Like low-severity forests in the western United States, many dry mixed-severity types experienced significant increases in stand density during the 20th century, threatening forest health and biodiversity, however not all understory development in mixed-severity forests increases the threat of severe wild fires. In general, current landscapes have been homogenized, reducing beta diversity and increasing the probability of large fires and insect outbreaks. Further loss of old, fire tolerant trees is of particular concern, but understory diversity has been reduced as well. High stand densities on relatively dry sites increase water use and therefore susceptibility to drought and insect outbreaks, exacerbating a trend of increasing regional drying. The need to restore beta diversity while protecting habitat for closed-forest specialists such as the northern spotted owl call for landscape-level approaches to ecological restoration.     

Prescribed fire effects on bark beetle activity and tree mortality in southwestern ponderosa pine forests

Prescribed fire is an important tool in the management of ponderosa pine (Pinus ponderosa Dougl. ex Laws.) forests, yet effects on bark beetle (Coleoptera: Curculionidae, Scolytinae) activity and tree mortality are poorly understood in the southwestern U.S. We compared bark beetle attacks and tree mortality between paired prescribed-burned and unburned stands at each of four sites in Arizona and New Mexico for three growing seasons after burning (2004–2006). Prescribed burns increased bark beetle attacks on ponderosa pine over the first three post-fire years from 1.5 to 13% of all trees, increased successful, lethal attacks on ponderosa pine from 0.4 to 7.6%, increased mortality of ponderosa pine from all causes from 0.6 to 8.4%, and increased mortality of all tree species with diameter at breast height >13 cm from 0.6 to 9.6%. On a per year basis, prescribed burns increased ponderosa pine mortality from 0.2% per year in unburned stands to 2.8% per year in burned stands. Mortality of ponderosa pine 3 years after burning was best described by a logistic regression model with total crown damage (crown scorch + crown consumption) and bark beetle attack rating (no, partial, or mass attack by bark beetles) as independent variables. Attacks by Dendroctonus spp. did not differ significantly over bole heights, whereas attacks by Ips spp. were greater on the upper bole compared with the lower bole. Three previously published logistic regression models of tree mortality, developed from fires in 1995–1996 in northern Arizona, were moderately successful in predicting broad patterns of tree mortality in our data. The influence of bark beetle attack rating on tree mortality was stronger for our data than for data from the 1995–1996 fires. Our results highlight canopy damage from fire as a strong and consistent predictor of post-fire mortality of ponderosa pine, and bark beetle attacks and bole char rating as less consistent predictors because of temporal variability in their relationship to mortality. The small increase in tree mortality and bark beetle attacks caused by prescribed burning should be acceptable to many forest managers and the public given the resulting reduction in surface fuel and risk of severe wildfire.     

Resilience of European larch (<Emphasis Type="Italic">Larix decidua</Emphasis> Mill.) forests to wildfires in the western Alps

European larch is a dominant species in the subalpine belt of the western Alps. Despite recent increases in wildfire activity in this region, fire ecology of European larch is poorly understood compared to other larch species around the world. This study aims to assess whether European larch forests are resilient to fires, and to find out the factors that drive such resilience. We assessed the recovery of larch forests along a gradient of fire severity (low, moderate, high) based on the abundance and dominance of post-fire larch regeneration. We established 200 plots distributed among burned larch forests in nine wildfires that occurred between 1973 and 2007 in the western Alps. We included variables regarding topography, climate, fire severity, fire legacies, ground cover, grazing intensity, and time since fire. To evaluate potential drivers of larch recruitment, we applied generalized linear mixed models (GLMM) and random forests (RF). Larch regeneration was much more abundant and dominant in the moderate- and high-severity fire classes than in the low-severity class. More than half of the plots in the moderate- and high-severity classes were classified as resilient, i.e., post-fire larch regeneration was enough to recover a larch stand. GLMM and RF produced complementary results: fire severity and legacies, such as snags, canopy cover and distance to seed source, were crucial factors explaining post-fire larch recruitment. This study shows that fire has a positive effect on larch regeneration, and we conclude that European larch forests are highly resilient to mixed-severity fires in the western Alps.     

Spatially modeling wildland fire severity in pine forests of Galicia,Spain

Fire is a major disturbance in forests and one of the most important carbon emissions sources, which contributes to climate change. Carbon emissions are directly correlated with the degree of organic matter consumption or fire severity. Gaining knowledge about the relative strength of the various explanatory variables is essential to mitigate its environmental impact. We tested an approach that combines wind modeling, light detection and ranging (LiDAR), remotely sensed vegetation indices and topography data for assessing the occurrence of high-severity fire using the random forests ensemble learning method. Data from four wildfires that occurred in Galicia (northwestern Spain) were used to exemplify the application of this approach. The models predicted high-severity occurrence with a classification accuracy ranging from 77 to 94%. High-severity fire occurred more frequently in areas of high simulated wind speed, and more pronouncedly, for cases reported as wind-driven fires. High severity also occurred more frequently in areas of high terrain roughness, on sunny slopes and in low canopy base height stands. This approach allowed predicting spatially explicit fire severity at a mean scale level (resolution of 25 m) with accuracy rates from 80 to 95%. This approach may be helpful for fire managers when delimiting and planning fuel treatments for severity mitigation or during fire suppression, and for post hoc case studies.     

Relationship between landscape structure and burn severity at the landscape and class levels in Samchuck, South Korea

Fuel treatments for reducing fire risk are necessarily tied to the landscape structure including forest composition and configuration. Thus understanding the relationships between landscape structure and burn severity is important for developing guidelines and management strategies for fire-resilient forests. The goal of this study was to investigate the relationship between landscape structure as described by spatial pattern metrics and burn severity at the landscape and class levels. In 2000, a mostly severe fire burned 16,210 ha of dense forest located in Samchuck on the east coast of the Korean peninsula. Spatial pattern metrics including patch density, largest patch index, mean shape index, area-weighted mean shape index, Euclidean nearest neighborhood distance, and Shannon's diversity index, as well as topographic characteristics of slope and elevation, were correlated with burn severity based on delta Normalized Burn Ratio (dNBR) assessments. Regression tree analysis was also carried out with the same variables to avoid spatial autocorrelation and to reveal the relative importance of variables to burn severity. The results of this study strongly suggest that both composition and configuration of the forest cover patches are closely tied to burn severity. In particular, both the correlation analysis and regression tree analysis indicated that the area of red pine tree forest cover was the most significant factor in explaining the variance of burn severity. Topography and spatial configuration of forest cover patches were also significantly related to burn severity. The heterogeneity of forests also had a significant influence on burn severity. To reduce fire risk and increase the fire resilience of forests, forest managers and agencies need to consider enhancing the heterogeneity of forests when implementing fuel treatment schemes. However, such fuel treatments for landscape structure may only be effective under moderate weather conditions.     

A comparison of soil properties after contemporary wildfire and fire suppression

Forests that were subject to frequent wildfires, such as ponderosa pine/Douglas-fir forests, had fire-return intervals of approximately 6–24 years. However, fire suppression over the last century has increased the fire-return interval by a factor of 5 in these forests, possibly resulting in changes to the soil. The objective of this study was to determine if soils of recently burned areas (representative of the natural fire-return interval) have different properties relative to soils in areas without recent fire. To assess this, recent low-intensity, lightning-caused, spot wildfire areas were located within fire-suppressed stands of ponderosa pine/Douglas-fir of the central, eastern Cascade Mountains of Washington State. Soil horizon depths were measured, and samples collected by major genetic horizons. Samples were analyzed for pH, C, N, C/N ratio, cation exchange capacity (CEC), base saturation (%BS), hydrophobicity and extractable P. Results show very little difference in soil properties between sites burned by low-severity fires and those areas left unburned. Such minimal changes, from these low-severity fires, in soil properties from fire suppression suggest there has also been little change in soil processes.     

Physiological responses of ponderosa pine in western Montana to thinning, prescribed fire and burning season

Low-elevation ponderosa pine (Pinus ponderosa Dougl. ex. Laws.) forests of the northern Rocky Mountains historically experienced frequent low-intensity fires that maintained open uneven-aged stands. A century of fire exclusion has contributed to denser ponderosa pine forests with greater competition for resources, higher tree stress and greater risk of insect attack and stand-destroying fire. Active management intended to restore a semblance of the more sustainable historic stand structure and composition includes selective thinning and prescribed fire. However, little is known about the relative effects of these management practices on the physiological performance of ponderosa pine. We measured soil water and nitrogen availability, physiological performance and wood radial increment of second growth ponderosa pine trees at the Lick Creek Experimental Site in the Bitterroot National Forest, Montana, 8 and 9 years after the application of four treatments: thinning only; thinning followed by prescribed fire in the spring; thinning followed by prescribed fire in the fall; and untreated controls. Volumetric soil water content and resin capsule ammonium did not differ among treatments. Resin capsule nitrate in the control treatment was similar to that in all other treatments, although burned treatments had lower nitrate relative to the thinned-only treatment. Trees of similar size and canopy condition in the three thinned treatments (with and without fire) displayed higher leaf-area-based photosynthetic rate, stomatal conductance and mid-morning leaf water potential in June and July, and higher wood radial increment relative to trees in control units. Specific leaf area, mass-based leaf nitrogen content and carbon isotope discrimination did not vary among treatments. Our results suggest that, despite minimal differences in soil resource availability, trees in managed units where basal area was reduced had improved gas exchange and growth compared with trees in unmanaged units. Prescribed fire (either in the spring or in the fall) in addition to thinning, had no measurable effect on the mid-term physiological performance and wood growth of second growth ponderosa pine.     

Evidence of mixed-severity fires in the foothills of the Rocky Mountains of west-central Alberta, Canada

This study presents new evidence of historic low-to-moderate-severity fires, intermixed with high-severity fires, in the foothills of the Rocky Mountains of west-central Alberta, Canada. High-severity fires that burned 120-300 years ago initiated even-aged cohorts of fast-growing lodgepole pine at each of the six study sites. Evidence of subsequent, low-to-moderate-severity fires included single and double fire scars on thin-barked lodgepole pine that were as small as 3.6 cm in diameter at the time of the fire, but survived. These low-to-moderate-severity fires resulted in structurally complex stands with a broad range of tree diameters and multiple cohorts of lodgepole pine, white and black spruce, and subalpine fir. At the site level, fire return intervals were variable, ranging from 29 to 167 years, but most were <80 years. Of the 9 years in which we documented low-to-moderate-severity fires, only the fires in 1889 and 1915 scarred trees at more than one site, indicating that these fires were small and had local effects. The new knowledge of historical, low-to-moderate-severity fires provided by this study has important implications for ecologically sustainable forest management. Although we recognize that further research needs to determine the extent of low-to-moderate-severity fires at the landscape scale, our results clearly indicate that a mixed-severity fires occurred at least locally. A broader range of silvicultural systems than is currently practiced would be consistent with historic forest dynamics.     

Spatial variability in stand structure and density-dependent mortality in newly established post-fire stands of a California closed-cone pine forest

Fire is an important process in California closed-cone pine forests; however spatial variability in post-fire stand dynamics of these forests is poorly understood. The 1995 Vision Fire in Point Reyes National Seashore burned over 5000 ha, initiating vigorous Pinus muricata (bishop pine) regeneration in areas that were forested prior to the fire but also serving as a catalyst for forest expansion into other locales. We examined the post-fire stand structure of P. muricata forest 14 years after fire in newly established stands where the forest has expanded across the burn landscape to determine the important factors driving variability in density, basal area, tree size, and mortality. Additionally, we estimated the self-thinning line at this point in stand development and compared the size-density relationship in this forest to the theorized (−1.605) log-log slope of Reineke’s Rule, which relates maximum stand density to average tree size. Following the fire, post-fire P. muricata density in the expanded forest ranged from 500 to 8900 live stems ha⁻¹ (median density = 1800 ha⁻¹). Post-fire tree density and basal area declined with increasing distance to individual pre-fire trees, but showed little variation with other environmental covariates. Self-thinning (density-dependent mortality) was observed in nearly all stands with post-fire density >1800 stems ha⁻¹, and post-fire P. muricata stands conformed to the size-density relationship predicted by Reineke’s Rule. This study demonstrates broad spatial variability in forest development following stand-replacing fires in California closed-cone pine forests, and highlights the importance of isolated pre-fire trees as drivers of stand establishment and development in serotinous conifers.     

The influence of mountain pine beetle outbreaks and drought on severe wildfires in northwestern Colorado and southern Wyoming: A look at the past century

Outbreaks of bark beetles and drought both lead to concerns about increased fire risk, but the relative importance of these two factors is the subject of much debate. We examined how mountain pine beetle (MPB) outbreaks and drought have contributed to the fire regime of lodgepole pine forests in northwestern Colorado and adjacent areas of southern Wyoming over the past century. We used dendroecological methods to reconstruct the pre-fire history of MPB outbreaks in twenty lodgepole pine stands that had burned between 1939 and 2006 and in 20 nearby lodgepole pine stands that were otherwise similar but that had not burned. Our data represent c. 80% of all large fires that had occurred in lodgepole pine forests in this study area over the past century. We also compared Palmer Drought Severity Index (PDSI) and actual evapotranspiration (AET) values between fire years and non-fire years. Burned stands were no more likely to have been affected by outbreak prior to fires than were nearby unburned stands. However, PDSI and AET values were both lower during fire years than during non-fire years. This work indicates that climate has been more important than outbreaks to the fire regime of lodgepole pine forests in this region over the past century. Indeed, we found no detectable increase in the occurrence of high-severity fires following MPB outbreaks. Dry conditions, rather than changes in fuels associated with outbreaks, appear to be most limiting to the occurrence of severe fires in these forests.     

Canopy fuel characteristics and potential crown fire behavior in Aleppo pine (Pinus halepensis Mill.) forests

Canopy fuel characteristics that influence the initiation and spread of crown fires were measured in representative Aleppo pine (Pinus halepensis Mill.) stands in Greece. Vertical distribution profiles of canopy fuel load, canopy base height and canopy bulk density are presented. Aleppo pine canopy fuels are characterized by low canopy base height (3.0–6.5 m), while available canopy fuel load (0.96–1.80 kg/m²) and canopy bulk density (0.09–0.22 kg/m³) values are similar to other conifers worldwide. Crown fire behavior (probability of crown fire initiation, crown fire type, rate of spread, fireline intensity and flame length) in Aleppo pine stands with various understory fuel types was simulated with the most updated crown fire models. The probability of crown fire initiation was high even under moderate burning conditions, mainly due to the low canopy base height and the heavy surface fuel load. Passive crown fires resulted mostly in uneven aged stands, while even aged stands gave high intensity active crown fires. Assessment of canopy fuel characteristics and potential crown fire behavior can be useful in fuel management and fire suppression planning.     

Estimation of fire severity by use of Landsat TM images and its relevance to vegetation and topography in the 2000 Samcheok forest fire

The severity of the 2000 Samcheok forest fire was classified by using Landsat TM images, and the effects of vegetation structures and topographic conditions on fire severity were analyzed. The estimated normalized difference vegetation index differences between the pre and post-fire Landsat TM images were used as the criteria in determining the levels of fire severity–low, moderate, and extreme. According to the results from fire severity estimation, of the 10,600 ha forest stands, 28% was severely damaged by crown fires, 38% was moderately damaged, and the remaining 34% was damaged slightly by surface fires. The overall accuracy of the fire severity classification was 83% (Kappa coefficient = 0.76). The results of χ ²-tests showed that fire severity differed significantly with the vegetation and topographic conditions as follows. The coniferous stands, compared with the mixed and broad-leaved, were more vulnerable to fire damage; the higher the slope of fire sites, the greater the fire damage; the south was the most vulnerable aspect; fire severity of coniferous forest stands increased with increasing elevation. However, in the study area it was found that fire severity of broad-leaved forest stands were negatively related to the elevation of the corresponding fire sites and affected more by vegetation conditions rather than by topographic conditions.     

Historic and current fire regimes in the Great Xing’an Mountains,northeastern China: Implications for long-term forest management

Understanding both historic and current fire regimes is indispensable to sustainable forest landscape management. In this paper, we use a spatially explicit landscape simulation model, LANDIS, to simulate historic and current fire regimes in the Great Xing’an Mountains, in northeastern China. We analyzed fire frequency, fire size, fire intensity, and spatial pattern of burnt patches. Our simulated results show that fire frequency under the current fire scenario is lower than under the historic fire scenario; total area burnt is larger with lower fire intensity under the historic fire scenario, and smaller with higher fire intensity under the current fire scenario. We also found most areas were burned by high intensity fires under the current fire scenario, but by low to moderate fires under the historic fire scenario. Burnt patches exhibit a different pattern between the two simulation scenarios. Large patches burnt by high intensity class fires dominate the landscape under the current fire scenario, and under historic fire scenario, patches burnt by low to moderate fire intensity fires have relatively larger size than those burnt by high intensity fires. Based on these simulated results, we suggest that prescribed burning or coarse woody debris reduction should be incorporated into forest management plans in this region, especially on north-facing slopes. Tree planting may be a better management option on these severely burned areas whereas prescribed burning after small area selective cutting, retaining dispersed seed trees, may be a sound forest management alternative in areas except for the severely burned patches.     

Pinus pinaster Ait. tree mortality following wildfire in Spain

Maritime pine (Pinus pinaster Ait.) is the tree species most affected by wildfire in the Iberian Peninsula. Prediction of the probability of fire-injured tree mortality is critical for management of burned areas, evaluation of the ecological and economic impact of wildfire and prescribed fire planning and application. Pine bark beetles (Scolytidae) frequently attack burned maritime pine stands and cause extensive post-fire mortality throughout the Iberian Peninsula. In the present study, maritime pine trees were monitored for three years following 14 wildfires in four ecotypes in Spain (11 fires in Galicia (Galician ecotype - NW Spain), one fire in Portillo (Meseta-Castellana ecotype - Central Spain), one fire in Rodenal (Rodenal ecotype - Central Spain), and one fire in Genalguacil (Sierra Bermeja ecotype - SW Spain)). Data on tree attributes, crown and bole injury, ground fire severity, Ips sp. presence and tree survival were obtained by examining 3085 trees. Logistic regression models for predicting the probability of delayed maritime pine mortality were developed by use of generalized estimated equations (GEE). An ample range of response to fire damage in mortality was evident among the four ecotypes and different models were fitted for each. The most important variables for predicting tree mortality were total crown volume damaged, presence of Ips sp. attack and cambium kill rating. The results highlight the extensive presence of Ips sp. in burned maritime pine forests and its importance in tree mortality process, the ample range of response of P. pinaster, in terms of post-fire mortality, as well as the need to develop site specific mortality models for the different ecotypes of this species following fire.     

Fire history and fire–climate relationships along a fire regime gradient in the Santa Fe Municipal Watershed, NM, USA

The Santa Fe municipal watershed provides up to 40% of the city's water and is at high risk of a stand-replacing fire that could threaten the water resource and cause severe ecological damage. Restoration and crown fire hazard reduction in the ponderosa pine (PP) forest is in progress, but the historic role of crown fire in the mixed-conifer/aspen (MC) and spruce-dominated forests is unknown but necessary to guide management here and in similar forests throughout the southwestern United States. The objective of our study was to use dendroecological techniques to reconstruct fire history and fire–climate relationships along an elevation, forest type, and fire regime gradient in the Santa Fe River watershed and provide historical ecological data to guide management. We combined systematic (gridded) sampling of forest age structure with targeted sampling of fire scars, tree-ring growth changes/injuries, and death dates to reconstruct fire occurrence and severity in the 7016 ha study area (elevation 2330–3650 m). Fire scars from 141 trees (at 41 plots) and age structure of 438 trees (from 26 transects) were used to reconstruct 110 unique fire years (1296–2008). The majority (79.0%) of fires burned during the late spring/early summer. Widespread fires that scarred more than 25% of the recording trees were more frequent in PP (mean fire interval (MFI)_25% = 20.8 years) compared to the MC forest (31.6 years). Only 24% of the fires in PP were recorded in the MC forest, but these accounted for a large percent of all MC fires (69%). Fire occurrence was associated with anomalously wet (and usually El Niño) years preceding anomalously dry (and usually La Niña) years both in PP and in the MC forest. Fire in the MC occurred during more severe drought (mean summer Palmer Drought Severity Index; PDSI = −2.59), compared to the adjacent PP forest (PDSI = −1.03). The last fire in the spruce forest (1685) was largely stand-replacing (1200 ha, 93% of sampled area), recorded as fire scars at 68% of plots throughout the MC and PP forests, and burned during a severe, regional drought (PDSI = −6.92). The drought–fire relationship reconstructed in all forest types suggests that if droughts become more frequent and severe, as predicted, the probability of large, severe fire occurrence will increase.     

Comparing the influence of site quality,stand age,fire and climate on aboveground tree production in Siberian Scots pine forests

Temporal patterns of stem and needle production and total aboveground net primary production (ANPP) were studied at the tree and stand level along four chronosequences of Siberian Scots pine (Pinus sylvestris L.) forests differing in site quality (poor lichen type or the more fertile Vaccinium type) and in frequency of surface fires (unburned, moderately burned (fire return interval of approximately 40 years), or heavily burned (fire return interval of approximately 25 years)). The maximum range of variability in aboveground production was quantified for: (1) possible long-term changes in site quality; (2) stand age; (3) non-stand-replacing, recurring surface fires; and (4) interannual climate variability. For (1) and (2), total ANPP was low in the lichen-type chronosequence, reached a maximum of 170 g C m(-2) year(-1) after 100 years and decreased to 100 g C m(-2) year(-1) in older stands. Maximum ANPP in the Vaccinium-type chronosequence was 340 g C m(-2) year(-1) and occurred earlier in the 53-year-old stand than in the other stands. Along the lichen-type chronosequences, peak ANPP was paralleled by maximum carbon allocation to stem growth. (3) In mature trees, damage by recurrent surface fires decreased stem growth by 17 +/- 19% over a 10-year period relative to pre-fire values. At longer timescales, ANPP was hardly affected by fire-related differences in mortality. (4) Needle- plus stem-NPP, reconstructed for a 3-year period, varied within a range of 15 g C m(-2) year(-1) in the lichen-type stands and 35 g C m(-2) year(-1) in the Vaccinium-type stands. For the same period, the coefficient of variance was higher for needle-NPP (20 +/- 10%) than for stem-NPP (12 +/- 7%). Needle- and stem-NPP did not covary in time. Most 30-year time series of stem-NPP at the tree level exhibited strong autocorrelation. In older trees, stem-NPP was positively correlated with growing season precipitation. Thus, the factors driving variability in ANPP ranked according to their maximum influence as: stand age (controlled by the frequency of stand-replacing fires) > site quality > growth depression because of surface fire damage approximately equal age-related reduction in ANPP > interannual variability approximately equal long-term effects of fire (stand density reduction). In lichen-type forests, we found that ANPP at the landscape level declined sharply when the interval between stand-replacing fires was less than 120 years, illustrating that fire strongly influences ANPP of boreal Scots pine forests.     

The influence of prescribed crown fire on lodgepole pine dwarf mistletoe (Arceuthobium americanum) populations 33 years post‐fire

Dwarf mistletoes (Arceuthobium spp.) are a group of obligate, hemiparasitic plants that infect numerous species in the Pinaceae in North America. Wildland fire is considered to be the primary natural agent influencing the population and distribution of dwarf mistletoes across landscapes. Based on this understanding, prescribed fire has been suggested as a potential method for dwarf mistletoe sanitation and control; however, experimental work has primarily focused on prescribed surface fire. In this study, we report long‐term impacts of three experimental crown fires on dwarf mistletoe severity in infested lodgepole pine stands in Colorado 33 years post‐fire. The three fires achieved tree mortality rates ranging from 20% to 100%. Our results suggested a significant negative relationship between the amount of fire‐caused tree mortality and future dwarf mistletoe severity. These findings supported the presumed natural role of fire in altering dwarf mistletoe populations, which perhaps exhibits a linear relationship between fire‐caused host tree mortality and future dwarf mistletoe severity.     

The effect of fire severity on first-year seedling establishment in a <Emphasis Type="Italic">Pinus canariensis</Emphasis> forest on Tenerife,Canary Islands

The Canarian pine (Pinus canariensis) exhibits a striking combination of high adult resistance to fire and intermediate serotiny. Hence, the study of its post-fire regeneration can support valuable new insights about functional adaptations to fire. Here, we analyse the first-year seedling establishment after fire in a P. canariensis forest on the northern slope of Tenerife, Canary Islands. The effects of fire severity and other explanatory variables on the seedling density recorded 9 months after fire were examined. We detected a clear unimodal relationship between seedling density and fire severity, with maximum regeneration associated with intermediate fire severity and no regeneration associated with very high crown damage. The results suggested that high severity fires may have caused the partial destruction of the aerial seed bank and/or the creation of unfavourable seedbed conditions for germination and seedling emergence. The density of large pine trees, reflecting seed availability, was the second most important factor explaining the distribution of seedlings. Cover of scorched needle litter on the ground correlated strongly and positively with pine seedling density and negatively with fire severity. The complete lack of regeneration at sites most strongly affected by fire does not represent a major threat for the stand recovery of the Canarian pine, due to the very high tree resistance to fire and the tremendous capacity of the Canarian pine to resprout after fire. The observed very high seedling densities at sites with intermediate fire impacts can probably be related to both the complete liberation of the seed bank (including seeds stored in serotinous cones) and favourable micro-environmental conditions for seed germination and seedling establishment.