Next Article in Journal
Statistical Comparison and Assessment of Four Fire Emissions Inventories for 2013 and a Large Wildfire in the Western United States
Previous Article in Journal
Parametric Analysis of a Steel Frame under Fire Loading Using Monte Carlo Simulation
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Fire
Volume 5
Issue 1
10.3390/fire5010026
fire-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Past Logging and Wildfire Increase above Ground Carbon Stock Losses from Subsequent Wildfire

by
Nicholas Wilson
1,2,* and
Ross Bradstock
1
1
Centre for Environmental Risk Management of Bushfires, University of Wollongong, Wollongong 2522, Australia
2
Fenner School of Environment and Society, The Australian National University, Canberra 2601, Australia
*
Author to whom correspondence should be addressed.
Fire 2022, 5(1), 26; https://0-doi-org.brum.beds.ac.uk/10.3390/fire5010026
Submission received: 4 January 2022 / Revised: 10 February 2022 / Accepted: 15 February 2022 / Published: 16 February 2022
Browse Figures
Review Reports Versions Notes

Abstract

:
Background: Wildfire is known to reduce forest carbon stocks, but the influence of antecedent disturbance on wildfire related carbon stock losses is not as well understood. Disturbances such as logging and wildfire may increase the vulnerability of remaining carbon stocks to subsequent wildfire. Conversely, these disturbances may reduce the impact of subsequent wildfire, resulting in lower carbon stock losses. Methods: We measured above ground carbon stocks in productive resprouting Eucalypt dominated forests before and after a mixed severity fire that burned during the 2019/20 ‘Black Summer’ fire season in south-eastern Australia. The initial surveys were stratified by time since logging and time since wildfire, allowing for an assessment of how these disturbance histories influence above ground carbon stock losses caused by subsequent wildfire. Results: Above ground carbon stock losses varied substantially; however, there was a weak decrease in losses associated with time since logging but not time since wildfire. Variance in carbon stock losses associated with logging were greater than that caused by the severity of the 2019/20 wildfire itself. Carbon losses and predicted effects of disturbance may be underestimated in some cases due to the accumulation of carbon at sites between pre- and post-fire surveys. Conclusions: This study presents the largest published dataset of direct carbon stock changes resulting from wildfire in eucalypt forests. Our findings indicate that logging reduces the stability of above ground carbon stocks in resprouting eucalypt forests. This information will be critical for land managers looking to manage forests for carbon sequestration.

1. Introduction

Natural disturbances such as wildfire are a major source of carbon loss from forest carbon pools [1,2,3,4] Some researchers and land managers are concerned that past land management may have reduced the stability of forest carbon pools, potentially converting them to net carbon sources [4,5,6]. Specifically, logging and altered fire regimes may have increased the prevalence of forests dominated by small trees that are more susceptible to fire induced mortality and have more fire sensitive carbon pools [6,7,8]. This subject is not well understood in relatively fire tolerant, resprouting Eucalypt forests [4] where there are few studies directly measuring pre- and post-wildfire carbon stocks, Volkova, et al. [9] being the one exception that the authors are aware of at the time of writing.
Above ground carbon is overwhelmingly stored in live trees [10], with large stocks of carbon sometimes present in the coarse woody debris of recently logged sites [11]. Carbon stocks may take decades to centuries to recover after logging but may recover within approximately 40 years after wildfire [11]. Wildfire generally causes carbon losses directly through combustion of finer fuel such as leaf litter and foliage) and indirectly through the mortality and decomposition of live vegetation such as trees and shrubs [12]. Because most carbon is stored in trees [13,14,15], mortality due to fire potentially causes higher losses of carbon in the long term, compared with immediate losses due to combustion [12]. The fire tolerance of the dominant tree species is also likely to influence fire related carbon stock losses. Forests dominated by resprouting Eucalypt species are less likely to experience high mortality rates after wildfire [16] due to their unique capacity to survive high severity fire by resprouting new foliage along the stem and branches [17]. This is reflected in smaller carbon stock losses after wildfire in resprouting, compared to non-resprouting Eucalypt forests [11]. Thus, resprouting Eucalypt forests, which can have significant carbon stocks [10], may be a relatively stable forest carbon pool [5]. However, carbon stock losses are likely to increase with fire severity, even in resprouting forests [8,18], due to greater mortality and consumption of living vegetation [7,16,19,20,21]
Some resprouting Eucalypt forests in south-eastern Australia may have a high prevalence of younger, smaller trees due to logging [10,22,23] and to a lesser extent wildfire [8,24]. Forests consisting of small trees may experience greater mortality as smaller trees have a reduced ability to successfully resprout after fire [8,19,25]. Additionally, short inter-fire intervals can erode the resprouting capacity of Eucalypts [26,27]. Individual trees may also be more vulnerable to subsequent fire due to damage caused by logging [23] and wildfire [28]. Thus, recent disturbance in resprouting Eucalypt forests, and the resultant prevalence of vulnerable trees, may reduce the stability of above ground carbon stocks [6]. This could result in greater above ground carbon stock losses from subsequent wildfire. On the other hand, recently disturbed forests, particularly recently logged forests, are likely to store less carbon [11], which may result in potential for greater stock losses from longer undisturbed forest (as per Keith, Lindenmayer, Mackey, Blair, Carter, McBurney, Okada and Konishi-Nagano [7]).
Vast areas (circa. 6 million hectares) of resprouting Eucalypt forest in south eastern Australia burnt during the 2019–2020 ‘Black Summer’ fires due to their unprecedented size [29,30]. Here we present analysis of the influence of time since logging and time since wildfire on the change in above ground carbon stocks caused by the 2019–2020 wildfires. Recent logging may have increased the potential for carbon stock losses owing to the prevalence of carbon in vulnerable pools such as small trees and coarse woody debris. On the other hand, recently logged sites may lose less carbon as they have a smaller stock to deplete. Recent fire may increase carbon stock losses due to the decreased resprouting capacity of recently burnt trees or may decrease carbon stock losses as vulnerable stocks were depleted during the previous fire. Regardless of disturbances history, we expected a positive relationship between above ground carbon stock losses and the severity of the 2019/20 fires.

2. Materials and Methods

2.1. Study Area and Survey Design

We measured above ground carbon stocks in forests with a known logging and wildfire history on the New South Wales South Coast (Figure 1). The forests are dominated by resprouting Eucalypt species (Corymbia maculata, Eucalyptus pilularis and Corymbia gummifera), with an understorey of emergent trees and tall shrubs and litter, low shrubs and grasses on the ground. The forests occur on the foothills of the coastal escarpment below approximately 400 m above sea level. Mean annual rainfall across the study area ranges from 943 to 1163 mm and mean annual temperatures from 15.4 to 16.2 °C. Selective logging in the region has occurred for over a century and expanded greatly in the 1950s [31]. Silvicultural practices vary but generally aim to retain a portion of smaller trees for future harvesting as well as habitat trees. Because post-harvest regeneration burns are not typically applied in this study area, logging slash (non-merchantable harvested tree material) remains in situ, forming a large carbon pool in recently logged sites [11]. Mixed severity wildfires occur at a multi-decadal frequency and are generally limited to drought periods that allow the abundant fuel to dry out [32]. Unfortunately, there was no reliable information on historic (i.e., prior to the 2019–2020 fires) logging intensity or wildfire severity for each site. In this study area, a site that has not experienced logging or wildfire for approximately 80 years loses 147 Mg of carbon per hectare after logging, compared to 15 Mg of carbon per hectare after wildfire [11].
Initial surveys of above ground carbon stocks were conducted between August 2018 and March 2020 (n = 90) [11]. The initial surveys were stratified by time since logging and wildfire to determine the independent effects of these two disturbances on above ground carbon stocks. Seventy-nine of these sites were sampled before November 2019, of which most subsequently burnt during the 2019/20 fire season by the Currowan and Clyde Mountain fires that started in late November and December 2019 respectively. The Currowan fire started from a lightning strike and later split to form the Clyde Mountain fire, cumulatively burning 598,437 ha over 82 days. Site selection was stratified to sample at least one site in every available combination of time since logging and time since wildfire at a decadal resolution (so that there was a known logging and wildfire history for each site). We sampled sites ranging from zero to approximately 80 years since logging and wildfire (Figure 1). Hazard reduction burning (planned fires to reduce fuel) had also been applied at some sites (Figure 1); however, we do explicitly stratify our sample by this disturbance type, as it was not expected to have a significant effect on total carbon stocks.
Post fire surveys of 59 sites were conducted between May and August 2020 using the methods in Wilson, Bradstock and Bedward [11]. The remaining 31 sites were not surveyed after the 2019–2020 fire for at least one of a number of reasons: (i) access was prohibited; (ii) they had been logged since the first survey; (iii) they had been affected by firefighting activities; (iv) it was not clear if they were fully burnt by the 2019–2020 fire. We did not identify a sufficient number of sites that did not burn during the 2019–2020 fires to provide an adequate sample of unburnt control sites. Given all sites were in the process of recovery after logging, we expect that there may have been small increases in carbon stocks across most sites that we are unable to account for due to the absence of unburnt control sites. Thus, our results may slightly underestimate losses at some sites. Above ground carbon was measured in a 20 × 50 m (0.1 ha) plots that were marked using a GPS with positional accuracy within 15 m. Positioning of the plot for the post fire survey was informed by GPS coordinates and initial site photos. We used features such as the number of large trees (relative to other trees on site)—which are easy to identify and unlikely to be completely removed from the site by a single fire—to assess if the plot was accurately positioned. We estimated above ground carbon from allometric equations for live and dead canopy forming trees, understorey vegetation, coarse woody debris and leaf litter.

2.2. Live and Dead Trees

The diameter at breast height (DBH) over bark of all live trees was measured in the 20 × 50 m plot (trees > 20 cm DBH) or a nested 10 × 50 m plot (5–20 cm DBH trees). All trees < 5 cm DBH in a nested 5 × 50 m plot were counted and attributed as 2.5 cm DBH. We did not account for variation in bark thickness associated with different species or bark consumption by fire, which is likely to have caused some variation in our estimates of tree biomass. We used the allometric equation described by Ximenes, et al. [34] (Table 1) to estimate the biomass of each tree. This equation was developed by weighing merchantable sized trees in our study area. We assumed 50% of the biomass was carbon for all live trees and shrubs. All dead standing trees greater than 10 cm DBH were recorded using the same stratified sampling technique used for live trees. We also recorded the proportion of each dead tree remaining to the nearest 25% and level of decay in three classes: 1-wood hard and intact (specific wood density of 0.78 g cm3, 47.81% carbon content); 2-decay extending to, but not within heartwood (specific wood density of 0.7 g cm3, 48.08% carbon content); 3-decay extending to the heartwood (specific wood density of 0.41 g cm 3.48% carbon content) [10]. Necromass in dead trees was estimated using the allometric equations for live trees in [34] and decay multipliers of 1, 0.897 and 0.526 for decay classes 1, 2 and 3 respectively. These decay multipliers reflect the change in specific wood density associated with each decay class. Carbon content was calculated using the assumed carbon content associated with each decay class (listed above).

2.3. Understorey Vegetation

The stem diameter at 10 cm above the ground of all living non-Eucalypt woody vegetation (referred to as understorey from here) were recorded in a nested 5 × 50 m plot. Stems < 2 cm diameter were not recorded. Biomass for each individual was estimated using the equation described by [35] (Table 1). This equation is not specific for our study area or the species within it. Given that our study area is in a relatively productive area compared to the that of Paul, Roxburgh, Chave, England, Zerihun, Specht, Lewis, Bennett, Baker and Adams [35], these equations may underestimate carbon in some cases. However, we expect understory carbon to make a relatively small contribution to total above ground carbon stocks.

2.4. Coarse Woody Debris

We measured the diameter and decay class of all coarse woody debris > 2.5 cm that intersected three 50 m transects positioned within each plot. Measurements were made at the point of interception. The carbon content of each piece was calculated by estimating the volume using the equation from Van Wagner [36] (Table 1) and multiplying it by the wood density and carbon content corresponding to the decay class described above. The equations Van Wagner [36] developed are universally applicable and the decay classes were developed for Eucalypt forests [10].

2.5. Litter

The depth of litter ≤ 2.5 cm diameter was measured at five equidistant points along the centre line of each plot. The cover of litter in the surrounding area was also recorded as greater or less than 50% from visual estimation. The biomass was estimated using one of two equations described McCarthy [37] (Table 1) that correspond to the cover of litter. Each litter mass estimate was multiplied by 0.5 to obtain the carbon content and averaged for each plot. This method was chosen over more accurate and resource intensive methods because the litter carbon pool was not the primary focus of the original study, nor was it expected to greatly affect the total carbon stock. This equation was developed in similar forests further south on the east coast of Australia.
We did not measure pyrogenic carbon because it was not the focus of the original research design and therefore not measured in the pre-fire survey. Consequently, it was not possible to compare the change in pyrogenic carbon between surveys.

2.6. Site Data

We recorded the latitude and longitude, the severity of the 2019/20 fire and the topographic position of each plot during the post fire survey. Fire severity (scorch and consumption of vegetation by the fire) was classified in the following four categories: extreme (>50% consumption of the canopy), high >90% canopy scorch, <50% canopy consumption), moderate (20–90% canopy scorch) and low (<10% canopy scorch, >10% scorch of the understorey), as per Gibson, et al. [38]. Due to the low number of sites burnt at extreme severity, we combined high and extreme severity sites into a single high severity category for analysis. Topographic position was classified as ridge, slope, flat or gully based on site inspection and assessment of topographic maps. The number of years since the last logging event, wildfire and hazard reduction fire (planned fires to reduce fuel) at the time of the post-fire survey were derived from mapping of the area of past logging and fire. We classified the time since hazard reduction as <10 years (approximately the duration of efficacy in affecting fire behaviour [39]), 10–35 years (all other mapped hazard reduction fires) and >35 years (no fire recorded within 35 years) since fire. The elevation, slope and aspect of each site were derived from a 30 m digital elevation model and mean annual temperature, mean annual rainfall, solar radiation and vapor pressure deficit were derived from 1 km WorldClim data [40].

2.7. Analysis

We used Bayesian regression models to assess the effect of antecedent logging, wildfire and hazard reduction and the severity of the 2019–2020 wildfire on above ground carbon stock losses associated with the 2019–2020 wildfire using the BRMS package (Bayesian Regression Models using ‘Stan’) [41] in R version 4.0.0 [42]. Our response variables were the change in above ground carbon stocks between the pre- and post-fire survey at each site for each of the six carbon pools (total, live tree, dead tree, understorey, coarse woody debris and litter carbon stocks). Total carbon was calculated as the sum of the live tree, dead tree, understorey, coarse woody debris and litter carbon stocks for a given survey. We considered modelling the proportional change in carbon stocks, but there was substantial variance in this data, which combined with the small sample size, made model predictions unreliable. Each response variable was predicted using a smoothed interaction between time since logging and time since wildfire (i.e., years since the last logging or wildfire event preceding the 2019–2020 wildfire) to account for interactions between these two disturbance types, time since last hazard reduction fire preceding the 2019–2020 wildfire) and the severity of the 2019–2020 wildfire. This enabled us to identify how disturbance history affects carbon stock losses after subsequent wildfire while accounting for the severity of the wildfire itself. We also fitted a smoothed interaction between the latitude and longitude of each site to account for spatial correlation, confounding effects of the spatial distribution of disturbance history and spatial patterns in carbon stocks identified by Wilson, Bradstock and Bedward [11]. Smooth terms are routinely fitted in ecological models to account for non-linear effects [43]. The residuals from each model were fitted against all remaining site data to identify if the inclusion of any other variables could improve the model fit; however, none were identified. The models were fitted using the ‘skew normal’ family and default priors with 5000 iterations, including 2500 warmup iterations, for each of four Markov chains. The ‘adapt_delta’ value was set to 0.999 and ‘max_treedpath’ value to 10 to assist with the model fitting. We assessed the model fit by comparing the prior and posterior response distributions and using Rhat values and trace plots of the Markov chains to check for convergence.

3. Results

We collected above ground carbon data in 59 plots before and after they were burnt during the 2019–2020 fire season. Of these 21 burnt at low severity, 10 at moderate severity and 28 at high severity (including 1 that was originally classified as extreme severity). Pre-fire above ground carbon stocks varied from 44.78 to 294.37 Mg ha−1 and post fire above ground carbon stocks varied from 22.87 to 250.41 Mg ha−1 (Figure 2). In both instances these were largely comprised of carbon stored in live trees and to a lesser extent coarse woody debris (Figure 2). Above ground carbon stocks changed between the pre- and post-fire surveys by 14.01 to −101.1 Mg ha−1 in total, 22.96 to −92.95 Mg ha−1 from live trees, 37.61 to −7.78 Mg ha−1 from dead trees, 0.45 to −16.41 Mg ha−1 from understorey vegetation, 33.99 to −80.83 Mg ha−1 from coarse woody debris and 0.41 to −11.54 Mg ha−1 from leaf litter (Figure 2). Proportionally, total and tree carbon stocks declined by less than a quarter in most sites between the pre- and post-fire surveys (Figure 2). While dead tree and coarse woody debris carbon stocks generally declined, there was substantial variance and some sites experienced large increases (for example, an 8394% increase in dead tree carbon and a 1176% increase in coarse woody debris carbon) (Figure 2). Understorey and litter carbon generally experienced declines close to 100%, but there were several sites with smaller losses and some with proportional increases (Figure 2).
Across all 59 plots, the median live tree stem density declined from 590 stems ha−1 in the pre-fire survey to 360 stems ha−1 in the post-fire survey (Figure 3). The median decline in stem density was 27%. Stems smaller than 10 cm DBH contributed most to the decline in stem density, although there were also notable declines in stems up to 30 cm DBH (Figure 3).

3.1. Disturbance History

Total above ground carbon losses declined with time since logging, although the effect was weak and there was substantial variance (Figure 4). The greater loss of carbon in recently logged sites was largely driven by the loss of coarse woody debris (Figure 4). Logging history appeared to have no effect on either live or dead tree carbon stock losses (Figure 4). Both understorey vegetation and litter carbon stock losses increased with time since logging (Figure 4). Total above ground carbon losses were not affected by wildfire history (Figure 4). Coarse woody debris carbon stocks losses decreased with time since wildfire but were effectively offset by increasing live tree carbon stock losses with time since wildfire (Figure 4).
Total above ground carbon and litter carbon stock losses were not affected by prior hazard reduction burning (Figure 5). Live tree carbon stock losses were lower at sites that had experienced hazard reduction burning in the previous 10 years compared to those that had not, while dead tree carbon stock losses were lower at sites that had not experienced a hazard reduction burn in the previous 35 years, compared to those that had (Figure 5). Understorey and coarse woody debris carbon stock losses tended to be greater where hazard reduction fire occurred <10 years prior to sampling compared to those that had not (Figure 5).

3.2. Fire Severity

Total carbon stock losses increased slightly with increasing severity of the 2019–2020 fire (Figure 6). The trend was largely the consequence of positive effect of severity on carbon losses from live trees and coarse woody debris (Figure 6). Moderate fire severity caused greater understorey carbon stock losses, while dead tree and litter carbon stock losses were not affected by fire severity (Figure 6).
Carbon stock losses increased from east to west in all pools except coarse woody debris, which declined from east to west (Figure S1). Total, live tree and dead tree carbon stock losses decreased from north to south, while coarse woody debris, understorey and litter stocks increased from north to south (Figure S1).

4. Discussion

This study is the largest direct comparison of pre- and post-fire above ground carbon stocks in resprouting Eucalypt forests. Our results demonstrated that disturbance (logging and fire) history had a modest influence on the change in above ground carbon stocks after wildfire in productive resprouting Eucalypt forests. These findings indicated that carbon stocks losses were greater at sites that had been recently logged, although losses were generally small compared to the total above ground stocks in mature forests. Losses were not clearly influenced by past wildfire but were greater at sites that burnt at higher fire severity during the 2019–2020 wildfires. Recent logging was associated with the loss of an additional 11.2 Mg ha−1 after the 2019–2020 wildfires compared to the longest unlogged (83 years since logging) and unburnt (81 years since wildfire) sites in our sample. By contrast, recent wildfire reduced losses by 3.3 Mg ha−1 after the 2019–2020 wildfire compared to the longest unlogged and unburnt sites. Logging history was a slightly greater source of variance in carbon stock losses from the 2019–2020 wildfire than the severity of the 2019–2020 wildfire, which resulted in 5.6 Mg ha−1 more carbon lost at sites burnt at a high severity than those burnt at low severity.
The high degree of variability we observed in above ground carbon stock losses was similar to that found in resprouting Eucalypt forests burnt by wildfire in Victoria [8]. We observed greater changes in above ground carbon stocks than other resprouting forests after wildfire [8,9] and hazard reduction burning [44]. This likely reflected the greater total above ground carbon stocks in our study area. We also note that our study is the only one to directly measure above ground carbon stocks at the same site before and after a wildfire, with the exception of [9] which only had five sites with repeat measurements. The carbon stock losses we observed were much lower than those observed after wildfire in more productive non-resprouting Eucalypt forests in Victoria [7].
The carbon stock loss values presented here are estimates that may not fully capture the changes in above ground carbon stocks after wildfire. For instance, we did not account for loss of foliage in live trees, which can form up to 6.8% of a tree’s biomass [45]. Although we note that most trees that we surveyed that had crown scorch or consumption were already resprouting new foliage prolifically. Some of the carbon lost between surveys will have transferred into pyrogenic carbon, which we did not measure. Carbon emissions from wildfires can be overestimated by 2 to 27% if pyrogenic carbon is not accounted for [46]. However, our focus is on the change in carbon stored on site above ground, not necessarily carbon emissions. Much of our study area received heavy rainfall in the months after the fire, which may have removed a considerable portion of pyrogenic carbon from our sites [47]. There may also have been some error in our estimates of carbon in coarse woody debris because we did not account for changed carbon concentrations associated with charring (see Aponte, et al. [48]). However, we suggest this error may be small given that the change in carbon within coarse woody debris was driven by the complete removal or addition of mainly large pieces of coarse woody debris, rather than small changes in diameter or decay class. The timing of our surveys may also have affected our estimates. Some surveys were performed 16 months prior to being burnt and appeared to have experienced small but measurable gains in carbon stored in live trees. The timing of our post fire surveys (three to eight months post fire) may also have influenced our classification of vegetation as dead or alive given the delays in resprouting after the fire. However, we assumed that these mechanisms only had small effects relative to the overall changes in above ground carbon stocks after wildfire.

4.1. Disturbance History

Despite having lower pre-fire above ground carbon stocks [11], recently logged sites experienced greater above ground carbon stock losses than longer unlogged sites. These changes were driven by the combustion of coarse woody debris in recently logged sites, rather than differences in stock losses from live trees. Large quantities of coarse woody debris are created during logging events in our study area. This pool accounted for much of the pre-fire carbon stock in recently logged forests but appears to have been substantially depleted by the 2019–2020 wildfire, resulting in greater total above ground carbon stock losses in recently logged sites compared to longer unlogged sites. The absence of an effect of logging on carbon stock losses from live trees may be due to the low quantity of carbon available to be lost in recently logged sites. Recently logged sites were dominated by smaller live trees that store little carbon, but these trees were more likely to be killed during the 2019–2020 fires (Figure 3). By contrast, the transfer of some exceptionally large live trees to the dead tree and coarse woody debris pools (Figure 3) resulted in highly variable losses, particularly from longer unlogged sites (Figure 4). While much carbon can be stored in dead trees, we noted many instances of recently fallen trees losing large portions of their biomass, and therefore carbon stocks, to combustion. These findings contrast with those of [7] who observed increasing carbon stock losses with stand age in non-resprouting forests. However, the forests studied by [7] are prone to stand replacement after high severity fire, and most coarse woody debris is consumed in post-logging regeneration burns [49]. This highlights the importance of considering the fire response type of dominant trees in Eucalypt forests and variations in silvicultural practices when assessing potential effects of subsequent wildfire.
Although total above ground carbon stocks were essentially unaffected by wildfire history, there were contrasting effects on live tree and coarse woody debris carbon stocks. Greater losses of coarse woody debris occurred in recently burnt sites. This pattern was also apparent after hazard reduction fire. A possible explanation was that the influx of carbon to the coarse woody debris pool-from vulnerable live trees, tree limbs or dead trees-caused by fire had already occurred in recently burnt sites. Thus, there was little vulnerable carbon to replace the coarse woody debris consumed by the subsequent 2019–2020 wildfire. The creation of a large pool of vulnerable dead carbon stocks was observed by Miesel, et al. [50] in North American conifer forests. By contrast, long unburnt sites may have contained more carbon that could be transferred into the coarse woody debris pool after fire. The weak increase in carbon loses from live trees with time since wildfire suggested that recent fire had already depleted vulnerable live tree carbon stocks, rather than increasing the prevalence of vulnerable live trees. The effect of past fire may have also been difficult to detect as we did not account for past fire severity, which may influence subsequent vulnerability [18,19]. Further, reductions in fire severity and associated carbon stock losses may only occur within approximately a decade after antecedent fire [27,39,51]. We may not have detected these effects in our data as we examined multi-decadal time scales.
Some of the variance in the data may be explained by variable logging intensity and wildfire severity prior to the 2019–2020 fires. Variation in the impacts of these two antecedent disturbances would likely influence the quantity and stability of carbon stocks that were burnt by the 2019–2020 fires. There are many variables and interactions to consider here, hence we do not speculate on how this may affect our results, other than to say that it is a likely source of unexplained variance.

4.2. Fire Severity

The increase in loss of carbon from low compared to high severity fire also may have been driven by losses from live trees and coarse woody debris, consistent with our predictions. Higher fire severity is known to cause greater mortality of live vegetation [19] and combustion of dead vegetation [52]. A notable result is that coarse woody debris tended to increase at sites that burnt at low and moderate severity but decreased at high severity. This effect was also observed by [52] in damp resprouting Eucalypt forest. This suggests that low and moderate fire severity created more coarse woody debris, from falling shrubs, live trees and their branches after the fire, than was consumed. By contrast, high severity fire may have resulted in the combustion of most material that could have entered the coarse woody debris pool. The absence of a clear effect of fire severity on live tree carbon is consistent with small differences in mortality after low and high severity fire observed in other resprouting Eucalypt forests [24]. We are unable to explain why moderate fire severity caused substantially greater losses of understorey carbon but noted that there were relatively few sites that burnt at moderate severity and pre-fire understorey carbon and post fire understorey carbon losses varied substantially.

4.3. Environmental Variation

The greater variance in carbon loss explained by latitude and longitude were likely associated with the quantity of carbon available to be lost. Analysis of the pre-fire data showed carbon stocks were highest in the eastern and northern most sites [11]. Thus, the absolute loss of carbon may have been proportional to the productivity and consequent size of the carbon stock available to be lost. While this explanation may be construed as contrary to the trend of lower carbon loss in long unlogged sites, lower carbon storage in recently logged sites was the likely consequence of the removal of large fire-resistant live trees, rather than lower site productivity.

4.4. Managing Carbon Stability

The results confirm that above ground carbon stocks in resprouting Eucalypt forests are relatively resistant to wildfire [5]. However, there were moderately higher losses of above ground carbon stocks from wildfire at recently logged sites compared to longer unlogged sites. This effect may have compounded the initial larger loss of carbon from logging itself [11]. That is, above ground carbon stocks were substantially reduced by logging and were further depleted i Effects of multiple fires on the carbon stability of fire-tolerant f burnt by wildfire shortly afterward. However, the greater risk may be in the frequency of fire in logged resprouting Eucalypt forests [15]. The carbon stocks of resprouting Eucalypt forests may become less stable if burnt too frequently [18]. This decline in stability may be more rapid in logged forests [15], which are typically dominated by smaller, more vulnerable live trees [8]. For example, live tree losses between surveys in our study were largely concentrated in smaller DBH classes (Figure 3). Further, fire severity may have been greater due to the potential for younger forests to be more flammable [53,54]. Thus, the interaction of logging with fire will be the key determinant of carbon stocks. Changes to silvicultural practices or harvesting intensity may have some capacity to mitigate against the modest increases in carbon losses associated with recent logging. However, our research was not designed to evaluate these alternatives and we cannot provide any recommendations for alternative practices informed by our data.
Hazard reduction burning has been recommended as a way of reducing carbon stock losses in the event of wildfire [9]. However, this was not apparent in our analyses or other similar research in resprouting Eucalypt forests [8]. Given the potential for frequent fire to destabilise carbon stocks, particularly in forests comprised of smaller live trees [8,15,18], hazard reduction burning may need to be applied cautiously in logged Eucalypt forests. However, we noted a weak reduction in live tree carbon losses associated with hazard reduction burning in the previous 10 years. Given that there were neither strong negative or positive effects of hazard reduction burning on in situ carbon stocks, any effects of burning programs may be largely indirect via potential changes to wildfire size and intensity.

5. Conclusions

Our study quantified variable losses of above ground carbon stocks due to wildfire in productive resprouting Eucalypt forests. Carbon stock losses were most strongly influenced by site productivity and the consequent quantity of carbon available to be lost. However, our results indicated that recent logging could increase above ground carbon stock losses. The effect of recent logging was greater than the variance associated with fire severity. Thus, the use of hazard reduction burning to reduce fire severity may not be able to directly counter the effects of logging on carbon losses from wildfire, although indirect mitigating effects on wildfire size and intensity may partially mitigate carbon losses. Indeed, we did not find strong evidence that hazard reduction burning is a useful tool for reducing carbon stock losses from wildfire. Given that carbon storage is a key objective of forest management, management of risks posed by logging and wildfire will be critical in maintaining carbon storage.

Supplementary Materials

The following supporting information can be downloaded at: https://0-www-mdpi-com.brum.beds.ac.uk/article/10.3390/fire5010026/s1, Figure S1: Marginal effects of variables predicting above ground carbon pools with 90% confidence intervals.

Author Contributions

Conceptualization, N.W. and R.B.; methodology, N.W. and R.B.; formal analysis, N.W.; investigation, N.W.; resources, N.W. and R.B.; data curation, N.W.; writing—original draft preparation, N.W.; writing—review and editing, N.W. and R.B.; visualization, N.W.; supervision, R.B.; project administration, N.W.; funding acquisition, N.W. and R.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Ecological Society of Australia (Student Research Award) and The Centre for Sustainable Ecosystem Solutions (Enhancement Fund).

Data Availability Statement

Data available from https://github.com/NickWilsonResearch/PostFireCarbon.

Acknowledgments

We thank Michael Bedward for advice on model specification, each of the volunteers who helped to collect data in the field, Christopher Gordon for advice on sampling techniques and Meaghan Jenkins and Fabiano Ximenes for advice on the selection of allometric equations. We also wish to thank Forestry Corporation of NSW for providing research permits, site access and mapping.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Norris, J.; Arnold, S.; Fairman, T. An indicative estimate of carbon stocks on Victoria’s publicly managed land using the FullCAM carbon accounting model. Aust. For. 2010, 73, 209–219. [Google Scholar] [CrossRef]
  2. Bradford, J.B.; Fraver, S.; Milo, A.M.; D’Amato, A.W.; Palik, B.; Shinneman, D.J. Effects of multiple interacting disturbances and salvage logging on forest carbon stocks. For. Ecol. Manag. 2012, 267, 209–214. [Google Scholar] [CrossRef]
  3. Bowman, D.M.; Murphy, B.P.; Neyland, D.L.; Williamson, G.J.; Prior, L.D. Abrupt fire regime change may cause landscape-wide loss of mature obligate seeder forests. Glob. Chang. Biol. 2014, 20, 1008–1015. [Google Scholar] [CrossRef] [PubMed]
  4. Bowman, D.M.; Williamson, G.J.; Price, O.F.; Ndalila, M.N.; Bradstock, R.A. Australian forests, megafires and the risk of dwindling carbon stocks. Plant Cell Environ. 2021, 44, 347–355. [Google Scholar] [CrossRef] [PubMed]
  5. Bowman, D.M.; Murphy, B.P.; Boer, M.M.; Bradstock, R.A.; Cary, G.J.; Cochrane, M.A.; Fensham, R.J.; Krawchuk, M.A.; Price, O.F.; Williams, R.J. Forest fire management, climate change, and the risk of catastrophic carbon losses. Front. Ecol. Environ. 2013, 11, 66–68. [Google Scholar] [CrossRef] [Green Version]
  6. Hurteau, M.D.; North, M.P.; Koch, G.W.; Hungate, B.A. Opinion: Managing for disturbance stabilizes forest carbon. Proc. Natl. Acad. Sci. USA 2019, 116, 10193–10195. [Google Scholar] [CrossRef] [Green Version]
  7. Keith, H.; Lindenmayer, D.B.; Mackey, B.G.; Blair, D.; Carter, L.; McBurney, L.; Okada, S.; Konishi-Nagano, T. Accounting for biomass carbon stock change due to wildfire in temperate forest landscapes in Australia. PLoS ONE 2014, 9, e107126. [Google Scholar] [CrossRef] [Green Version]
  8. Bennett, L.T.; Bruce, M.J.; Machunter, J.; Kohout, M.; Krishnaraj, S.J.; Aponte, C. Assessing fire impacts on the carbon stability of fire-tolerant forests. Ecol. Appl. 2017, 27, 2497–2513. [Google Scholar] [CrossRef]
  9. Volkova, L.; Meyer, C.M.; Murphy, S.; Fairman, T.; Reisen, F.; Weston, C. Fuel reduction burning mitigates wildfire effects on forest carbon and greenhouse gas emission. Int. J. Wildland Fire 2014, 23, 771–780. [Google Scholar] [CrossRef]
  10. Roxburgh, S.; Wood, S.; Mackey, B.; Woldendorp, G.; Gibbons, P. Assessing the carbon sequestration potential of managed forests: A case study from temperate Australia. J. Appl. Ecol. 2006, 43, 1149–1159. [Google Scholar] [CrossRef]
  11. Wilson, N.; Bradstock, R.; Bedward, M. Comparing forest carbon stock losses between logging and wildfire in forests with contrasting responses to fire. For. Ecol. Manag. 2021, 481, 118701. [Google Scholar] [CrossRef]
  12. Keith, H.; Lindenmayer, D.; Mackey, B.; Blair, D.; Carter, L.; McBurney, L.; Okada, S.; Konishi-Nagano, T. Managing temperate forests for carbon storage: Impacts of logging versus forest protection on carbon stocks. Ecosphere 2014, 5, 1–34. [Google Scholar] [CrossRef] [Green Version]
  13. Bennett, L.T.; Aponte, C.; Baker, T.G.; Tolhurst, K.G. Evaluating long-term effects of prescribed fire regimes on carbon stocks in a temperate eucalypt forest. For. Ecol. Manag. 2014, 328, 219–228. [Google Scholar] [CrossRef]
  14. Fedrigo, M.; Kasel, S.; Bennett, L.T.; Roxburgh, S.H.; Nitschke, C.R. Carbon stocks in temperate forests of south-eastern Australia reflect large tree distribution and edaphic conditions. For. Ecol. Manag. 2014, 334, 129–143. [Google Scholar] [CrossRef]
  15. Collins, L.; Bradstock, R.; Ximenes, F.; Horsey, B.; Sawyer, R.; Penman, T. Aboveground forest carbon shows different responses to fire frequency in harvested and unharvested forests. Ecol. Appl. 2019, 29, e01815. [Google Scholar] [CrossRef] [Green Version]
  16. Benyon, R.G.; Lane, P.N. Ground and satellite-based assessments of wet eucalypt forest survival and regeneration for predicting long-term hydrological responses to a large wildfire. For. Ecol. Manag. 2013, 294, 197–207. [Google Scholar] [CrossRef]
  17. Burrows, G. Buds, bushfires and resprouting in the eucalypts. Aust. J. Bot. 2013, 61, 331–349. [Google Scholar] [CrossRef]
  18. Bennett, L.T.; Fairman, T.A.; Nitschke, C.R.; Aponte, C. Effects of multiple fires on the carbon stability of fire-tolerant eucalypt forests depends on fire frequency and severity. In Proceedings of the 2019 ESA Annual Meeting, ESA, Washington, DC, USA, 11–16 August 2019. [Google Scholar]
  19. Bennett, L.T.; Bruce, M.J.; MacHunter, J.; Kohout, M.; Tanase, M.A.; Aponte, C. Mortality and recruitment of fire-tolerant eucalypts as influenced by wildfire severity and recent prescribed fire. For. Ecol. Manag. 2016, 380, 107–117. [Google Scholar] [CrossRef]
  20. Dunn, C.J.; Bailey, J.D. Tree mortality and structural change following mixed-severity fire in Pseudotsuga forests of Oregon’s western Cascades, USA. For. Ecol. Manag. 2016, 365, 107–118. [Google Scholar] [CrossRef]
  21. Etchells, H.; O’Donnell, A.J.; McCaw, W.L.; Grierson, P.F. Fire severity impacts on tree mortality and post-fire recruitment in tall eucalypt forests of southwest Australia. For. Ecol. Manag. 2020, 459, 117850. [Google Scholar] [CrossRef]
  22. Benson, J. The effect of 200 years of European settlement on the vegetation and flora of New South Wales. Cunninghamia 1991, 2, 343–370. [Google Scholar]
  23. Watson, G.M.; French, K.; Collins, L. Timber harvest and frequent prescribed burning interact to affect the demography of Eucalypt species. For. Ecol. Manag. 2020, 475, 118463. [Google Scholar] [CrossRef]
  24. Vivian, L.M.; Cary, G.J.; Bradstock, R.A.; Gill, A.M. Influence of fire severity on the regeneration, recruitment and distribution of eucalypts in the Cotter River Catchment, Australian Capital Territory. Austral Ecol. 2008, 33, 55–67. [Google Scholar] [CrossRef]
  25. Collins, L. Eucalypt forests dominated by epicormic resprouters are resilient to repeated canopy fires. J. Ecol. 2020, 108, 310–324. [Google Scholar] [CrossRef]
  26. Fairman, T.A.; Bennett, L.T.; Tupper, S.; Nitschke, C.R. Frequent wildfires erode tree persistence and alter stand structure and initial composition of a fire-tolerant sub-alpine forest. J. Veg. Sci. 2017, 28, 1151–1165. [Google Scholar] [CrossRef]
  27. Fairman, T.A.; Bennett, L.T.; Nitschke, C.R. Short-interval wildfires increase likelihood of resprouting failure in fire-tolerant trees. J. Environ. Manag. 2019, 231, 59–65. [Google Scholar] [CrossRef] [PubMed]
  28. Bowman, D.; Kirkpatrick, J. Establishment, suppression and growth of Eucalyptus delegatensis RT Baker in multiaged forests. I. The effects of fire on mortality and seedling establishment. Aust. J. Bot. 1986, 34, 63–72. [Google Scholar] [CrossRef]
  29. Boer, M.M.; de Dios, V.R.; Bradstock, R.A. Unprecedented burn area of Australian mega forest fires. Nat. Clim. Chang. 2020, 10, 171–172. [Google Scholar] [CrossRef]
  30. Bowman, D.; Williamson, G.; Yebra, M.; Lizundia-Loiola, J.; Pettinari, M.L.; Shah, S.; Bradstock, R.; Chuvieco, E. Wildfires: Australia needs national monitoring agency. Nature 2020, 584, 188–191. [Google Scholar] [CrossRef] [PubMed]
  31. Kearney, D. Native Forest Silviculture Manual; Foresty Cprporation: Beecroft, Australia, 2015. [Google Scholar]
  32. Nolan, R.H.; Boer, M.M.; Resco de Dios, V.; Caccamo, G.; Bradstock, R.A. Large-scale, dynamic transformations in fuel moisture drive wildfire activity across southeastern Australia. Geophys. Res. Lett. 2016, 43, 4229–4238. [Google Scholar] [CrossRef] [Green Version]
  33. State Government of NSW Department of Planning‚ Industry and Environment. Fire Extent and Severity Mapping (FESM). 2020. Available online: https://www.planning.nsw.gov.au/ (accessed on 3 January 2022).
  34. Ximenes, F.A.; Gardner, W.D.; Richards, G.P. Total above-ground biomass and biomass in commercial logs following the harvest of spotted gum (Corymbia maculata) forests of SE NSW. Aust. For. 2006, 69, 213–222. [Google Scholar] [CrossRef]
  35. Paul, K.I.; Roxburgh, S.H.; Chave, J.; England, J.R.; Zerihun, A.; Specht, A.; Lewis, T.; Bennett, L.T.; Baker, T.G.; Adams, M.A. Testing the generality of above-ground biomass allometry across plant functional types at the continent scale. Glob. Chang. Biol. 2016, 22, 2106–2124. [Google Scholar] [CrossRef] [PubMed]
  36. Van Wagner, C. The line intersect method in forest fuel sampling. For. Sci. 1968, 14, 20–26. [Google Scholar]
  37. McCarthy, G.J. Report 44: Surface Fine Fuel Hazard Rating—Forest Fuels in East Gippsland; Department of Sustainability and Environment: East Melbourne, Australia, 2004; p. 33. [Google Scholar]
  38. Gibson, R.; Danaher, T.; Hehir, W.; Collins, L. A remote sensing approach to mapping fire severity in south-eastern Australia using sentinel 2 and random forest. Remote Sens. Environ. 2020, 240, 111702. [Google Scholar] [CrossRef]
  39. Tolhurst, K.G.; McCarthy, G. Effect of prescribed burning on wildfire severity: A landscape-scale case study from the 2003 fires in Victoria. Aust. For. 2016, 79, 1–14. [Google Scholar] [CrossRef]
  40. Fick, S.E.; Hijmans, R.J. WorldClim 2: New 1-km spatial resolution climate surfaces for global land areas. Int. J. Climatol. 2017, 37, 4302–4315. [Google Scholar] [CrossRef]
  41. Bürkner, P.-C. brms: An R package for Bayesian multilevel models using Stan. J. Stat. Softw. 2017, 80, 1–28. [Google Scholar] [CrossRef] [Green Version]
  42. R Core Team. R: A Language and Environment for Statistical Computing, 4.0.0; R Foundation for Statistical Computing: Vienna, Austria, 2020. [Google Scholar]
  43. Pedersen, E.J.; Miller, D.L.; Simpson, G.L.; Ross, N. Hierarchical generalized additive models in ecology: An introduction with mgcv. PeerJ 2019, 7, e6876. [Google Scholar] [CrossRef] [Green Version]
  44. Volkova, L.; Weston, C.J. Carbon loss from planned fires in southeastern Australian dry Eucalyptus forests. For. Ecol. Manag. 2015, 336, 91–98. [Google Scholar] [CrossRef]
  45. Applegate, G.B. Biomass of Blackbutt (Eucalyptus pilularis Sm.) Forests on Fraser Island; University of New England: Armidale, Australia, 1982. [Google Scholar]
  46. Santin, C.; Doerr, S.H.; Jones, M.W.; Merino, A.; Warneke, C.; Roberts, J.M. The relevance of pyrogenic carbon for carbon budgets from fires: Insights from the FIREX experiment. Glob. Biogeochem. Cycles 2020, 34, e2020GB006647. [Google Scholar] [CrossRef]
  47. Rumpel, C.; Ba, A.; Darboux, F.; Chaplot, V.; Planchon, O. Erosion budget and process selectivity of black carbon at meter scale. Geoderma 2009, 154, 131–137. [Google Scholar] [CrossRef]
  48. Aponte, C.; Tolhurst, K.G.; Bennett, L.T. Repeated prescribed fires decrease stocks and change attributes of coarse woody debris in a temperate eucalypt forest. Ecol. Appl. 2014, 24, 976–989. [Google Scholar] [CrossRef]
  49. VicForests. VicForests 2015 Ecologically Sustainable Forest Management Plan; VicForests: Melbourne, Australia, 2015. [Google Scholar]
  50. Miesel, J.; Reiner, A.; Ewell, C.; Maestrini, B.; Dickinson, M. Quantifying changes in total and pyrogenic carbon stocks across fire severity gradients using active wildfire incidents. Front. Earth Sci. 2018, 6, 41. [Google Scholar] [CrossRef] [Green Version]
  51. Barker, J.W.; Price, O.F. Positive severity feedback between consecutive fires in dry eucalypt forests of southern Australia. Ecosphere 2018, 9, e02110. [Google Scholar] [CrossRef]
  52. Bassett, M.; Chia, E.K.; Leonard, S.W.; Nimmo, D.G.; Holland, G.J.; Ritchie, E.G.; Clarke, M.F.; Bennett, A.F. The effects of topographic variation and the fire regime on coarse woody debris: Insights from a large wildfire. For. Ecol. Manag. 2015, 340, 126–134. [Google Scholar] [CrossRef]
  53. Price, O.F.; Bradstock, R.A. The efficacy of fuel treatment in mitigating property loss during wildfires: Insights from analysis of the severity of the catastrophic fires in 2009 in Victoria, Australia. J. Environ. Manag. 2012, 113, 146–157. [Google Scholar] [CrossRef]
  54. Taylor, C.; McCarthy, M.A.; Lindenmayer, D.B. Nonlinear effects of stand age on fire severity. Conserv. Lett. 2014, 7, 355–370. [Google Scholar] [CrossRef] [Green Version]
Fire 05 00026 g001 550
Figure 1. Location of the study area in south-eastern Australia (left), the individual plots within it and the severity of the 2019–2020 wildfires (right) (a). Severity mapping sourced from State Government of NSW Department of Planning‚ Industry and Environment [33]. Time since the last logging event, wildfire and hazard reduction fire prior to the 2019–2020 wildfires at the time of the post fire surveys for all 59 sites (b).
Figure 1. Location of the study area in south-eastern Australia (left), the individual plots within it and the severity of the 2019–2020 wildfires (right) (a). Severity mapping sourced from State Government of NSW Department of Planning‚ Industry and Environment [33]. Time since the last logging event, wildfire and hazard reduction fire prior to the 2019–2020 wildfires at the time of the post fire surveys for all 59 sites (b).
Fire 05 00026 g001
Fire 05 00026 g002 550
Figure 2. Distribution of measured above ground carbon stocks from the pre- and post-fire surveys (a), the change in above ground carbon stocks between the two surveys (b) and the proportional change in above ground carbon stocks between the two surveys in each carbon pool (c). Distributions are represented by the median (centre line of the box), the 25th and 75th percentile (lower and upper limits of the box respectively), the 2.5th and 97.5th percentiles (lower and upper limits of the whiskers respectively) and outliers (individual points). Nine outliers from the dead tree pool and seven outliers from the coarse woody debris pool were removed as they exceeded 200% and made visualization of the remaining data difficult.
Figure 2. Distribution of measured above ground carbon stocks from the pre- and post-fire surveys (a), the change in above ground carbon stocks between the two surveys (b) and the proportional change in above ground carbon stocks between the two surveys in each carbon pool (c). Distributions are represented by the median (centre line of the box), the 25th and 75th percentile (lower and upper limits of the box respectively), the 2.5th and 97.5th percentiles (lower and upper limits of the whiskers respectively) and outliers (individual points). Nine outliers from the dead tree pool and seven outliers from the coarse woody debris pool were removed as they exceeded 200% and made visualization of the remaining data difficult.
Fire 05 00026 g002
Fire 05 00026 g003 550
Figure 3. Distribution of live tree stem densities in 10 cm DBH classes from the pre- and post-fire surveys. Each box and whisker represents the variation in live tree stem abundance across all 59 plots for the given survey and stem class. Distributions are represented by the median (centre line of the box), the 25th and 75th percentile (lower and upper limits of the box respectively) and the 2.
Figure 3. Distribution of live tree stem densities in 10 cm DBH classes from the pre- and post-fire surveys. Each box and whisker represents the variation in live tree stem abundance across all 59 plots for the given survey and stem class. Distributions are represented by the median (centre line of the box), the 25th and 75th percentile (lower and upper limits of the box respectively) and the 2.
Fire 05 00026 g003
Fire 05 00026 g004 550
Figure 4. Change in above ground carbon stocks between the pre- and post-fire carbon surveys plotted against time since logging and time since the previous wildfire for total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks. in each of the carbon pools. Solid lines and transparent ribbons indicate the median and 90% posterior density intervals respectively fitted by three levels of time since wildfire with observed values (black dots). Predictions were made with fire severity fixed at ‘high’, topographic position at ‘Slope’, time since hazard reduction at 10–35 years and all other variables at their mean.
Figure 4. Change in above ground carbon stocks between the pre- and post-fire carbon surveys plotted against time since logging and time since the previous wildfire for total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks. in each of the carbon pools. Solid lines and transparent ribbons indicate the median and 90% posterior density intervals respectively fitted by three levels of time since wildfire with observed values (black dots). Predictions were made with fire severity fixed at ‘high’, topographic position at ‘Slope’, time since hazard reduction at 10–35 years and all other variables at their mean.
Fire 05 00026 g004
Fire 05 00026 g005 550
Figure 5. Change in total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks between the pre- and post-fire carbon surveys in response to time since the last hazard reduction fire. Central point and bounds indicate the median and 90% posterior density intervals respectively. Predictions were made with time since logging and time since fire at 40 years, fire severity fixed at ‘high’, topographic position at ‘Slope’ and all other variables at their mean. CWD = coarse woody debris.
Figure 5. Change in total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks between the pre- and post-fire carbon surveys in response to time since the last hazard reduction fire. Central point and bounds indicate the median and 90% posterior density intervals respectively. Predictions were made with time since logging and time since fire at 40 years, fire severity fixed at ‘high’, topographic position at ‘Slope’ and all other variables at their mean. CWD = coarse woody debris.
Fire 05 00026 g005
Fire 05 00026 g006 550
Figure 6. Change in total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks between the pre- and post-fire carbon surveys in response to the severity of the 2019–2020 wildfire. Central point and bounds indicate the median and 90% posterior density intervals respectively. Predictions were made with time since logging and time since fire fixed at 40 years, topographic position at ‘Slope’, time since hazard reduction at 10–35 years and all other variables at their mean.
Figure 6. Change in total (a), live tree (b), dead tree (c), understorey (d), coarse woody debris (e) and litter (f) carbon stocks between the pre- and post-fire carbon surveys in response to the severity of the 2019–2020 wildfire. Central point and bounds indicate the median and 90% posterior density intervals respectively. Predictions were made with time since logging and time since fire fixed at 40 years, topographic position at ‘Slope’, time since hazard reduction at 10–35 years and all other variables at their mean.
Fire 05 00026 g006
Table
Table 1. Equations used to estimate biomass in each carbon pool and the sources they were derived from.
Table 1. Equations used to estimate biomass in each carbon pool and the sources they were derived from.
Carbon ComponentMeasured VariableSource
Trees (Live and dead)AGB (kg) = 57.6 − 12(DBH) + 0.92(DBH)2[34]
UnderstoreyAGB (kg) = exp (−3.007 + 2.428 × ln(D10)) × 1.128[35]
Coarse woody debrisVolume = π2∑Diameter2/(8 × Transect Length)[36]
Litter (high cover)Litter mass (kg) = (0.36 × Depth) − 1.21[37]
Litter (low cover)Litter mass (kg) = (0.41 × Depth) − 2.35[37]
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Wilson, N.; Bradstock, R. Past Logging and Wildfire Increase above Ground Carbon Stock Losses from Subsequent Wildfire. Fire 2022, 5, 26. https://0-doi-org.brum.beds.ac.uk/10.3390/fire5010026

AMA Style

Wilson N, Bradstock R. Past Logging and Wildfire Increase above Ground Carbon Stock Losses from Subsequent Wildfire. Fire. 2022; 5(1):26. https://0-doi-org.brum.beds.ac.uk/10.3390/fire5010026

Chicago/Turabian Style

Wilson, Nicholas, and Ross Bradstock. 2022. "Past Logging and Wildfire Increase above Ground Carbon Stock Losses from Subsequent Wildfire" Fire 5, no. 1: 26. https://0-doi-org.brum.beds.ac.uk/10.3390/fire5010026

Article Metrics

Back to TopTop