Fuel moisture in Mountain Ash forests with contrasting fire histories

Forest Ecology and Management 400 (2017) 568–577

Contents lists available at ScienceDirect

Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco

Fuel moisture in Mountain Ash forests with contrasting fire histories Jane G. Cawson a,⇑, Thomas J. Duff a, Kevin G. Tolhurst b, Craig C. Baillie a, Trent D. Penman b a b

School of Ecosystem and Forest Sciences, University of Melbourne, Burnley Campus, 500 Yarra Boulevard, Richmond, VIC 3121, Australia School of Ecosystem and Forest Sciences, University of Melbourne, Creswick Campus, Water Street, Creswick, VIC 3363, Australia

a r t i c l e

i n f o

Article history: Received 16 May 2017 Received in revised form 19 June 2017 Accepted 20 June 2017

Keywords: Canopy cover Fire severity Flammability Fuel availability Microclimate Plant area index Regime shift Time since fire Wildfire

a b s t r a c t Fuel moisture is a key driver of forest flammability as it influences ignition likelihood, fire intensity and resultant fire severity. Changes to forest canopy cover following disturbances like wildfire or logging may alter forest flammability by changing the microclimatic conditions that influence fine fuel moisture. Wet forests dominated by Mountain Ash (Eucalyptus regnans) are highly valued for their flora and fauna, timber, carbon and water. Wildfires are an important part of the lifecycle of these forests, but too frequent fire can threaten post-fire regeneration. With large tracts of Mountain Ash forest recovering from recent wildfires (in 2009 and 1983) there is a need to understand the mechanisms driving flammability in these forests, particularly as the forest structure changes following fire. This study sought to understand the effects of fire history on the flammability of Mountain Ash forests by specifically considering fuel moisture for different times since fire and fire severities. We measured canopy cover (plant area index) and fuel moisture within 8 forest sites last burnt between 7 and 200 years ago by wildfires of low or high severity. Fuel moisture and fuel availability (i.e. number of days when fine fuels are dry enough to ignite and sustain spreading fire) were strongly associated with canopy cover; with denser canopied forests having higher fuel moisture. The largest differences in canopy cover occurred between the recently burnt high and low severity forests. For the longer-unburnt forests there were no systematic differences evident in canopy cover with time since fire or fire severity. The fuel moisture was higher and fuels only available to burn on one day in the forest recently burnt by high severity fire (in association with a dense canopy). In contrast, fuels were drier and available to burn on 238 days in the forest recently burnt by low severity fire (in association with a sparser canopy). For the longer-unburnt forests (33 or more years since fire) there were no clear trends between fuel moisture and time since fire and fire severity suggesting that fires do not have a lasting impact on fuel moisture within these wet forests. Overall, this study shows that wildfires have immediate impacts on fuel moisture in Mountain Ash forests but as the time since fire increases, moisture appears to be more a function of canopy properties than fire history. Ó 2017 Elsevier B.V. All rights reserved.

1. Introduction Fire is important to many ecosystems globally, influencing the distribution, abundance and structural form of particular plant species and vegetation communities (Ryan, 2002; Bond and Keeley, 2005). Whilst being ecologically beneficial in many situations, extreme fire behaviour can also threaten human lives and property (Keeley et al., 2009; Koutsias et al., 2012; Blanchi et al., 2014) and altered fire regimes may pose risks to the ecosystems themselves (Bond and Keeley, 2005; Meyn et al., 2007). Flammability – the ability of vegetation to burn (Gill and Zylstra, 2005; Pausas et al., 2017) – is important to fire behaviour and fire regimes. To successfully manage fire, we need to understand the ⇑ Corresponding author. E-mail address: [email protected] (J.G. Cawson). http://dx.doi.org/10.1016/j.foreco.2017.06.046 0378-1127/Ó 2017 Elsevier B.V. All rights reserved.

factors driving forest flammability and how they vary spatially and temporally across the landscape. Fine fuel moisture is a key driver of flammability (Gill and Zylstra, 2005). Fine fuel is defined here as being the fuel (live and dead vegetation) burning in the flaming zone of a fire, typically dead fuel less than 6 mm thick and live fuel less than 2 mm thick (Tolhurst and Cheney, 1999). Fine fuel moisture contents influence the amount of fuel available to burn (n.b. in this study fuel availability is measured by the number of days when the fine fuels are dry enough to ignite and sustain a spreading fire) and this has implications for ignition likelihood, rate of spread, intensity and resultant severity of the fire (e.g. Rothermel, 1972; Fosberg et al., 1981; Forestry Canada Fire Danger Group, 1992; Burrows, 1999). For lower fine fuel moisture contents there is a higher likelihood of successful ignition, higher rates of spread and more intense fire behaviour. For Australian eucalypt forests, the upper

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

moisture threshold for ignition of dead fine fuel is likely to be 22%, while most prescribed burns occur up to a threshold of 16% and erratic fire behaviour occurs below 7% (McArthur, 1968; Sneeuwjagt and Peet, 1985; Sullivan et al., 2012). Over longer time scales, fuel moisture contents influence the frequency (or recurrence interval) of fire, with drier forests available to burn more often if there is sufficient biomass (Meyn et al., 2007; Bradstock, 2010). The level of moisture in the landscape also influences the total biomass or fuel load (Meyn et al., 2007; Bradstock, 2010). Wetter forests tend to have higher fuel loads (Thomas et al., 2014), resulting in intense fire behaviour when the fuel is dry enough to burn. An important consideration in relation to fine fuel moisture is how it varies spatially and temporally across landscapes. Fine fuel moisture contents are primarily controlled by microclimatic conditions – surface temperatures, relative humidity, precipitation and solar radiation (Countryman, 1977; Viney, 1991; Matthews, 2014). Mountain topography and forest cover create spatially variable microclimatic conditions over short distances (e.g. 100 m to 1 km) (Chen et al., 1999; Aussenac, 2000; Beniston, 2006). Several studies have measured fine fuel moisture contents in contrasting topographic positions and vegetation types (Matthews, 2014; Nyman et al., 2015). Those studies demonstrated substantial spatial variation in moisture content with typically higher moisture contents on polar-facing slopes, in gullies and beneath denser canopies compared with equatorial-facing slopes, ridges and sparser canopies. The spatial connectivity of available fuels within a catchment varies at a range of timescales. At hourly to daily timescales, this is likely to contribute to spatially heterogeneous fire behaviour and fire severity (e.g. Bradstock et al., 2010; Leonard et al., 2014). Drought conditions can cause landscapes to become highly connected as polar-facing slopes, gullies and denser canopied forests dry out, increasing the risk of major conflagrations (Miller and Urban, 2000; Caccamo et al., 2012; Sullivan and Matthews, 2013). Forest canopy cover often varies concurrently with topography, but can also vary independently in response to disturbances such a wildfire and logging. Canopies typically have a moderating effect on the climate by intercepting a portion of solar radiation and precipitation, reducing wind speeds and buffering against the extremes of temperature and potential evapotranspiration (Chen et al., 1999; Aussenac, 2000). In tropical forests selective logging and fire can reduce the density of the canopy, resulting in lower fuel moistures (Fetcher et al., 1985; Uhl and Kauffman, 1990; Holdsworth and Uhl, 1997; Balch et al., 2008). When coupled with changes to the fuel load and structure, the lower fuel moisture contents can make the forest more fire prone, leading to a fire regime shift with more frequent fires (Uhl and Kauffman, 1990; Balch et al., 2015). In the eucalypt forests of southern Australia it is common practice for harvested stands to be burnt adjacent to unharvested forest, relying on the differential in fuel moisture content between the closed standing forest and the open, harvested forest to prevent the fire from igniting the unharvested forest (Department of Conservation and Environment, 1990). Mountain Ash (Eucalytpus regnans) dominated wet forests are valued for their unique flora and fauna, timber, carbon and water, as well as being the tallest hardwood forests globally. Infrequent and intense wildfires are part of their lifecycle; the trees are killed and regenerate from seed, resulting in even-aged stands (Ashton, 1976). Lower intensity, understorey fires may also occur, but are less common (Ashton, 2000; Lindenmayer et al., 2000; Lindenmayer, 2009). Mountain Ash are obligate seeders requiring 25–30 years to reach maturity, as a result they are vulnerable to frequent fire, which can eliminate them from a site (Lindenmayer et al., 2011; Bowman et al., 2014). This has occurred in recent years in the higher elevation Alpine ash forests (E. delegatensis) (Bowman

569

et al., 2014; Bassett et al., 2015). Little is known about how the flammability of these wet forests changes as a result of fire. It has been suggested that fires and logging make the forest more fire prone and flammability declines with time since fire (Lindenmayer et al., 2009, 2011). However, the mechanisms driving this have not been studied in wet eucalypt forests. A number of factors could be important to consider, including fuel structure, fuel load, species composition and fuel moisture. Our study sought to better understand the effects of fire history on fuel moisture contents and fuel availability in Mountain Ash forests. We considered the effect of time since fire and past fire severity using automated fuel moisture sensors installed in forests with different fire histories. Our key research questions were:  To what extent does canopy cover drive differences in fine fuel moisture?  To what extent does fire history determine canopy cover?  Is there an association between fire history and fine fuel moisture?  What are the implications for forest flammability? 2. Methods 2.1. Site description Eight forested study sites were located in Mountain Ash forest (Eucalyptus regnans) in the Central Highlands region of Victoria, Australia (Fig. 1, Table 1). These forests occur in areas with high, relatively reliable rainfall (1000–1500 mm yr 1) and deep, fertile soils (Ashton and Attiwill, 1994). An additional two sites on flat clearings in close proximity to the forested sites were used for equivalent open weather observations. The forest sites spanned a range of major wildfire years (2009, 1983, 1939 and long unburnt) and fire severities. It is likely that the long unburnt ‘rainforest’ site had not been burnt by a high severity fire for >200 years (though a low severity fire may have occurred). It contained a dense midstorey of mature rainforest species (e.g. Nothofagus cuninghamii and Atherosperma moschatum) beneath an overstorey of mature and senescing Mountain Ash trees. The four Mountain Ash trees within the plot had diameters of 79, 186, 229 and 317 cm. Based on those diameters the estimated ages of the trees was 72, 168, 206, and 283 years old, respectively (Ashton, 1976). Sites were selected to be as similar as possible in all physical attributes except fire history, but some variation was unavoidable due to the geographic configuration of the past fires. All sites were approximately south-facing, in mid-hillslope positions with slopes between 11 and 23 and elevations between 373 and 740 m. The vegetation overstorey consisted of Mountain Ash, however structure and understorey species composition varied markedly between the sites. Our sites captured the broad vegetation successional stages described in ecological studies for Mountain Ash forests (Ashton, 2000). The Rainforest site occurred in a narrow strip along a subsurface drainage line in close proximity to the 1939High and 1939Low sites (same elevation and aspect as the 1939 sites). It is likely that the drainage line within the Rainforest site and the southerly aspect provided enough moisture and protection from fire to enable the rainforest to establish in this location (Wood et al., 2011). 2.2. Field data collection Fuel moisture contents were monitored for 16 months from November 2015 to March 2017, spanning two fire seasons. The fire season for Mountain Ash forests typically occurs in January and February (Murphy et al., 2013), though conditions are infrequently dry enough for the forests to burn even at this time of year. Fuel

570

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

Fig. 1. Location of forest and open weather stations within the Central Highlands of Victoria in south-eastern Australia. Shading depicts major wildfire zones for 1939, 1983 and 2009 (where wildfires overlapped, only the most recent fire area is shown). Numbers indicate the station identification and can be linked with the station names in Table 1.

Table 1 Physical site descriptors for the weather stations located in forest and open sites.

a b c d e

No.

Sitea

Elevation m

Rainfallb mm y 1

Stems ha 1

Mapped years since last firec

Tree diam cm

Age classd & percent of trees

Fire historye

1 2 3 4

2009Low 2009High 1983Low 1983High

37.3612, 37.3667, 37.4734, 37.4918,

145.3625 145.4122 145.4359 145.4213

13 12 9 11

240 178 248 212

373 580 689 678

1125 1550 1395 1385

108 1176 102 151

7 7 33 33

166

723

1580

18

77

37.4232, 145.4257

10

215

712

1580

71

77

1939High#2

37.4042, 145.3442

19

182

550

1550

101

77

8

Rainforest

37.4231, 145.4395

24

190

715

1580

38

Unknown

9 10

Open1 Open2

37.4248, 145.3344 37.3620, 145.4140

2 5

229 18

662 518

1275 1550

0 0

NA NA

77 y, 100% 7 y, 100% 77 y, 100% 33 y, 70% 77 y, 30% 77 y, 50% >200 y, 50% 77 y, 93% >200 y, 7% 77 y, 90% >200 y, 10% 77 y, 25% >200 y, 75% NA NA

Low severity in 2009 High severity in 2009 Low severity in 1983 High severity in 1983

11

77(37) 13(4) 64(27) 28(7) 68(10) 80 319 76 (18) 180 67(27) 214 79 244(67) NA NA

5

1939Low

37.4231, 145.4358

6

1939High

7

Coordinates

Slope

o

Aspect

o

Low severity in 1939 High severity in 1939 High severity in 1939 Low severity in 1939 NA, Open site NA, Open site

The site name describes the last fire year and the estimated fire severity. Derived from gridded climate statistics from the Australian Bureau Meteorology. Derived from Department of Environment Land Water and Planning (2016a). Calculated using tree diameters and equations from Ashton (1976). Based on mapped fire history and estimates of tree age based on size.

moistures were measured using a single Campbell Scientific 10-h fuel moisture stick at each site, installed at a height of 0.3 m above the ground. Fuel moisture measurements were recorded at 10-min intervals by a Campbell Scientific logger. The fuel moisture sticks were assumed to be indicative of fine fuel moisture contents. This assumption was verified by comparing the fuel stick moisture contents with manual fine fuel moisture contents. The fine fuel moisture contents were measured gravimetrically by collecting samples of surface (<10 mm deep), subsurface

(>10 mm deep) and elevated litter. Sampling occurred on four separate occasions in November 2015, February 2016, January 2017 and March 2017. Samples were collected in the field in water tight containers and oven dried at 105 °C for 48 h. On each occasion five samples were collected and averaged per site. Those averages were compared with the fuel stick moisture content observations from the same time of day (10-min fuel stick measurements were averaged over one hour). Pearson’s correlation coefficients for surface (0–10 mm), dead elevated and profile (combined surface and

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

subsurface litter) fine fuels were 0.87, 0.82 and 0.84, respectively (Fig. A.1). Those correlations between the sensors and fine fuel moisture contents were strong for all fuel types but surface and profile moisture contents were consistently higher than the fuel stick measurements. The fuel stick measurements most closely matched the dead elevated fine fuel moisture contents, which makes sense in terms of the placement of the sensor (30 cm above the forest floor). Time since last fire was determined from mapped last burnt records (Department of Environment Land Water and Planning, 2016a) for the sites burnt in 2009, 1983 and 1939 (Table 1). For the longest unburnt sites, we used mapped old growth (Department of Environment Land Water and Planning, 2016b) to locate them initially and then tree diameters to estimate tree and age and time since fire based on equations provided by Ashton (1976). Tree diameters were measured for the 12 closest trees to a plot point near the weather station (for sparser sites we only included trees within a 25 m radius). Severity was estimated from the time since last fire and tree age. Where the tree ages suggested that >70% trees had regenerated following the last fire, then the severity was classified as high. Otherwise the severity was classified as low. Similar approaches were used by Ough (2001) and Lindenmayer (2009). Plant Area Index (PAI) was calculated from hemispherical photos taken under overcast sky at two heights (0.3 and 1.2 m) immediately adjacent to the radiation sensors. PAI is half the total leaf and stem area per unit ground area. We used a Coolpix8400 camera with FC-E9 fisheye adapter and analysed the photos using HemiView Canopy Analysis Software (Delta-T-Devices), which employs a threshold technique to classify the images.

2.3. Data processing and analysis All data processing and analysis were performed using the R statistical programming language, Version 3.3.1 (R Core Team, 2016). To check for data errors we visually inspected the data in time-series graphs and removed data spikes and troughs associated with sensor failures; there were very few sensor failures during the study. PAI was summarised by calculating means and standard deviations for each site for measurements at 0.3 and 1.2 m. The fuel moisture data were summarised at daily time intervals by calculating daily minimums. Mean minimum fuel moisture contents were derived by sub-setting the minimum daily data based on time periods of interest (i.e. full study duration or peak fire season) and then calculating means over those timeframes. The minimum daily data from the open stations were averaged. Then, the daily minimum forest fuel moisture was subtracted from the averaged open fuel moisture to obtain a daily forest-to-open difference. Daily fuel availability (the number of days when the fine fuels are dry enough to ignite and sustain a spreading fire) was estimated from the minimum daily moisture. Only days with fuel moisture contents less than 16% were considered available based on research identifying this as the upper limit for prescribed burning in dry eucalypt forests (Sneeuwjagt and Peet, 1985; Tolhurst and Cheney, 1999; Slijepcevic et al., 2015). This is also the minimum moisture content allowed in an adjacent unharvested forest during coupe burning operations in Mountain Ash forest to prevent the spread of fire into the unharvested forest (Department of Conservation and Environment, 1990). It is important to note that fuel availability in this study is only a relative measure between sites not an absolute measure. In reality, fuel availability is likely to be less than the estimates provided in this paper because the fuel sticks were drier than the surface fine fuels (Fig. A.1) for which these threshold moisture conditions were designed. Furthermore,

571

the availability is calculated based on the minimum moisture content for the day, which may have occurred for less than one hour. The data were displayed using boxplots and time-series charts. Linear regressions were used to examine the association between PAI, fuel moisture and fuel availability. Wilcoxon rank sum tests were used to test the statistical significance of differences between means. A non-parametric test was chosen because the data were typically skewed. Pearson correlation coefficients were calculated to quantify the level of agreement between the fuel stick measurements and manual fine fuel moisture measurements.

3. Results Canopy cover – represented by Plant Area Index (PAI) – was strongly related to fuel moisture and fuel availability. Fig. 2 shows linear regressions between PAI and (a) fuel moisture throughout the study, (b) fuel moisture in the peak fire season, (c) fuel availability throughout the study. The fuel moisture contents represented here are average differences in the daily minimum fuel moisture between the open and forest sites. All the regressions were reasonably strong (R2 = 0.67, 0.69 and 0.75 for (a), (b) and (c) respectively) suggesting that PAI is important to fuel moisture and fuel availability. Boxplots in Fig. 3 illustrate the degree of variability in PAI both within and between the wildfire sites. Often the variability in PAI within a site was greater than the variation between sites. Differences between the sites were more apparent when PAI was measured at 0.3 m. At this height, PAI is a function of the tree canopy and shrub (or near surface) vegetation. 2009Low site had a significantly lower mean PAI at 0.3 m than the other sites combined (Wilcoxon rank sum test: p < 0.001). 2009High had a significantly higher PAI at 0.3 m than the other sites combined (Wilcoxon rank sum test: p = 0.011) but when sites were compared individually 2009High and 1983Low were not statistically significantly different (Wilcoxon rank sum test: p = 0.548). The largest difference in mean PAI at 0.3 m was between 2009Low (2.27) and 2009High (4.63) (Fig. 4). For the remaining sites, mean PAI at 0.3 m ranged from 3.1 to 4.5 and did not appear to vary incrementally as a function of time since fire or fire severity. The greatest difference in fuel moisture contents occurred between the 2009Low and 2009High wildfire sites (Table 2). Average minimums were 3 times lower for 2009Low (20.9%) compared with 2009High (61.6%) for the full study duration. That difference was slightly less during the peak fire season with average minimum moisture contents 2.5 times lower for 2009Low compared with 2009High. Fuel moisture contents were also relatively high at 1983Low, particularly during the peak fire season when they averaged 34.3%. For the remaining sites, there was less site-tosite variability in fuel moisture with average minimums ranging from 31.8 to 43.4% and 20.9 to 25.3% for the full study duration and peak fire season respectively. The boxplots in Fig. 5 show differences in minimum fuel moisture between open and forest sites calculated on a daily basis. The fuel moisture at 2009Low differed least from the open sites while the fuel moisture at 2009High and 1983Low differed most from the open sites. Mean values from individual forest sites were all significantly different from the mean of the other sites combined (Wilcoxon rank sum test: p < 0.05), with the exception of 1939Low during the peak fire season. However, some sites were clearly more similar to each other than others (e.g. Low2009 was clearly drier than the other sites whereas Rainforest and 1939High#2 had very similar mean values). The estimated number of days when the fine fuels were available to burn differed widely between the sites (Table 2 and Fig. 6). Fuels were available to burn for 238 days in 2009Low com-

572

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

Fig. 2. Simple linear regressions to predict (A) the mean difference in moisture content between open and forested sites as a function of PAI for the full time-series of data, (B) the mean difference in moisture content between open and forested sites as a function of PAI for the peak fire season (January and February) and (C) the relative number of days when the fuel was available to burn for the full time-series of data.

Fig. 3. Boxplots illustrating the range of values for plant area index (PAI) at each forest site measured at heights of (A) 0.3 m and (B) 1.2 m. Solid dots denote the mean PAI. PAI was estimated in five different locations at each site using hemispherical photography.

Fig. 4. Association between fire history (time since fire and severity) and PAI measured at heights of (A) 0.3 m and (B) 1.2 m. The PAI values are means derived from five (0.3 m) or six (1.2 m) photos taken surrounding weather station.

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

573

Table 2 PAI, fuel moisture and fuel availability statistics for full duration of study (November 2015 to March 2017) and peak fire season (January and February). Values reported are means; standard deviations in brackets for PAI. Shaded bars show magnitude of value relative to other sites.

Fig. 5. Summary of differences in minimum daily moisture content between open and forest sites for (A) full duration of study, (B) the peak fire season (January and February). Solid dots denote the mean. Moisture contents > 70% were excluded. The full study duration was from November 2015 to March 2017.

pared with only one day for 1983Low and 2009High. Fuel availability for the remaining sites ranged from 27 days (1939High) to 120 days (Rainforest). 4. Discussion 4.1. Relationship between canopy cover and fine fuel moisture The density of the canopy, as determined by PAI, was a strong predictor of differences in fuel moisture between the open and forest sites. The fuels in the sparser canopied forests with lower PAI had lower fuel moisture contents and were more frequently available to burn. A number of studies in tropical forests report a similar positive association between canopy cover and fuel moisture (Fetcher et al., 1985; Uhl and Kauffman, 1990; Holdsworth and Uhl, 1997; Ray et al., 2005). In contrast, for conifer forests the relationship between canopy cover and fuel moisture is less consistent. Some studies report lower fuel moisture in association with reduced canopy cover (Tanskanen et al., 2005; Ma et al., 2010) while others report little effect of canopy (Faiella and Bailey, 2007; Estes et al., 2012).

The canopy is thought to buffer the forest floor from extremes in temperature (Chen et al., 1999), primarily by regulating the amount of solar radiation reaching the forest floor (Countryman, 1977; Chen et al., 1999; Aussenac, 2000; Zou et al., 2007; von Arx et al., 2013). As a result, higher temperatures and lower relative humidity are typically observed beneath sparser canopied forests, leading to faster rates of fuel drying and lower fuel moisture contents (Countryman, 1977; Holdsworth and Uhl, 1997; Ray et al., 2005). Other mechanisms include rainfall interception and reduction in wind speeds, which are also a function of canopy density (Aussenac, 2000) and are important to fuel moisture (Matthews et al., 2007). Lower rates of rainfall interception beneath sparser canopied forests may cause higher fuel moisture contents immediately following rain although this effect is temporary because the sparse canopy also allows for faster drying of the fuel (Ray et al., 2005). Tree height and canopy density determine air movement within the forest with taller, denser forests having less air movement (Aussenac, 2000) and thus potentially lower evaporation and fuel drying rates (Ray et al., 2005). Macroclimatic conditions may be an important determinant of the degree to which canopy cover influences fuel moisture. Our

574

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

Fig. 6. Cumulative time-series of the number of days when the fuels were drier than 16% moisture content for each site.

study extended over two relatively wet fire seasons, potentially heightening the association between canopy cover and fuel moisture. In contrast, very dry climatic conditions may lead to a situation where fuels are dry irrespective of the canopy (Faiella and Bailey, 2007; Estes et al., 2012). A limitation of our study was that there was only one fuel moisture stick per site, which meant the spatial variability in fuel moisture contents within a site was not captured. Such variability is likely occur as a result of spatially variable solar radiation beneath the canopy (Countryman, 1977). Atypical results from the 1983Low site compared with the other sites that were >33 years since fire may be attributable to a dense but patchy layer of ferns that were 0.5 m in height. With so much ground cover close to the hemispherical camera lens, PAI at 0.3 m was relatively high on average (mean = 4.23) but also highly variable (standard deviation = 1.14). In the location of the fuel moisture stick the PAI at 0.3 m was particularly high (5.35), which probably meant that the fuel moisture contents recorded at this point were higher than the average conditions across the site. As a result the site has relatively high fuel moisture and low fuel availability compared with the other sites that were >33 years since fire. Future studies should use more than one fuel stick, especially for sites with dense near surface fuels where PAI is spatially variable. 4.2. Relationship between fire history and canopy cover There was a large difference in canopy cover (i.e. PAI) between the high and low severity recently burnt sites. This likely reflects differing ecological responses of the forests to high compared with low severity fire (Ashton, 1981; Ashton and Attiwill, 1994). For Mountain Ash forests impacted by high severity fires the majority of trees are killed and regenerate from seed, resulting in dense even-aged stands (Ashton, 1976). Stem densities can be very high in the early years post-fire; though the stem density at 2009High (1176 stems ha 1) was lower than densities reported elsewhere (Ashton, 1976). Additionally, mid-storey and understorey species regenerate densely with consistently high crown cover resulting in very dense vegetation cover overall (Attiwill, 1994; Ashton, 2000). For 2009High the PAI at 1.2 m was lower on average (mean = 4.52) than the PAI at 0.3 m (mean = 4.63), which reflects

the influence of ground vegetation (e.g. Tetrarhena juncea and Cassinia aculeata) on PAI closer to the forest floor. For Mountain Ash forests impacted by low severity fire, only some of the overstorey trees are killed (Benyon and Lane, 2013) potentially resulting in a multi-age forest (McCarthy and Lindenmayer, 1998; Lindenmayer et al., 2000). Few studies document the early stages of recovery from a low severity fire in Mountain Ash forest (one exception is Benyon and Lane, 2013). Our observations suggest that while much of the overstorey remains intact the midstorey and understorey are impacted. As a result, PAI may be less in a forest recovering from a low severity fire than in an unburnt mature forest. For 2009Low there were some dead trees in the canopy and mid-storey that were probably killed during the fire. The remaining mid-storey consisted predominately of Bedfordia arborescens (Blanket leaf) and Cyathea australis (Rough tree fern) with relatively sparse ground vegetation. As a result, PAI at 1.2 m (mean = 2.42) was higher than PAI at 0.3 m (mean = 2.27) because the camera lens was closer to the vegetation for PAI at 1.2 m. For the remaining sites, where fires occurred from 33 to 200 years ago, canopy cover did not appear to vary incrementally as a function of time since fire or fire severity. Other studies report changes to canopy cover for individual vegetation strata and species over time (Ashton, 1976, 2000). However, those changes in cover for individual vegetation strata and species do not necessarily equate to changes in the total canopy cover. Indeed, the total canopy cover is likely to reflect site productivity and therefore remain relatively constant over-time (Long et al., 2004; Long and Vacchiano, 2014), except in the immediate aftermath of a fire or disturbance when the canopy cover deviates from long-term equilibrium levels. Our results show that total canopy cover (combined understorey and tree canopy) recovers to equilibrium levels within 33 of a fire regardless of fire severity. Even during the early stages of regeneration at the high severity site (2009High), average canopy cover was only marginally higher than long-term equilibrium levels. 4.3. Association between fire history and fine fuel moisture The results show that young Mountain Ash forests are typically less available to burn than older forests. The fuels were below threshold levels for flammability for one day for 2009High compared with up to 238 days for the other sites. The high mean PAI for this site explains its higher fuel moisture. In contrast, fuel moistures were generally quite similar for stands 33–200 years old, reflecting similarities in canopy cover for forests within this age range (one exception was 1983Low as discussed earlier). Consequently, stand age does not appear to be a key driver of fuel moistures for forests older than 33 years. Fire severity was related to fuel moistures in the recently burnt forests with fuel moistures at 2009Low averaging three times lower than fuel moistures in 2009High. Additionally, fuels were available on 237 more days in 2009Low compared with 2009High. These results are associated with differences in canopy cover with 2009Low having significantly less canopy than 2009High. Similarly, fuel moistures in 2009Low were significantly lower than fuel moistures in the mature forest, also reflecting lower canopy in 2009Low compared with the other sites. A limitation of this study was that there were only 8 forest sites, which means caution should be exercised when using the results to make broad-scale predictions about the relationship between fire history and fuel moisture. Although the study sites were selected to be as similar as possible in all physical attributes except fire history, some variation was unavoidable due to the geographic configuration of the past fires. Those variations in site-specific factors (e.g. topography, soils, rainfall and past fires) could have contributed to differences in fuel moisture between the sites. Small

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

sample sizes are typical of studies like ours, where the costs associated with infrastructure purchase and maintenance are high. Consequently, the findings need to be considered in the context of similar research. 4.4. Implications for forest flammability Our results suggest that Mountain Ash forests in the first decade following high severity fire are typically less available to burn than longer unburnt forests. This could mean that the younger forests are less flammable. Certainly other studies report a lower incidence of fire and lower fire severity in newly regenerating wet eucalypt forests dominated by Mountain Ash and Karri (E. diversicolor) (Attiwill et al., 2014; Taylor et al., 2014). A more substantial difference in fuel moistures and fuel availability was seen when comparing the recently burnt low severity site (2009Low) to the other sites. Fuels were vastly more available at 2009Low and this suggests that low severity fires may increase the flammability of Mountain Ash forests. The impact of low severity fire in Mountain Ash forests seems to be analogous to the impact of selective logging in tropical rainforests. In both instances the total canopy cover is reduced and this is associated with lower fuel moistures and more available fuel. For tropical forests those lower fuel moistures have contributed to more frequent fires that may lead to a regime shift provided the fuel loads are sufficient to sustain repeated fires (Uhl and Kauffman, 1990; Balch et al., 2015). For Mountain Ash forests, low severity fires are a relatively unusual phenomenon, at least in recent decades. To date there is no quantitative evidence to show that they lead to more frequent fire. For the longer unburnt sites (33+ years) there were differences in fuel moisture and fuel availability but those differences did not occur incrementally as a function of time since fire or fire severity. Therefore, it appears that any changes to flammability caused by fire are short-lived in relation to the lifecycle of the Mountain Ash forest. Interestingly, the Rainforest site was not substantially wetter than the other sites, which raises the question of why rainforest patches are less flammable if indeed fuel moisture contents are not higher. In drawing conclusions relating to flammability a limitation with the study is that only fuel moisture was considered. Fuel moisture is only one (albeit important) factor that contributes to the flammability of a forest. Other factors include fuel load, fuel structure, floristic composition, ignition source, fire weather and seasonal dryness (Rothermel, 1972; Fosberg et al., 1981; Forestry Canada Fire Danger Group, 1992; Gould et al., 2007). Specifically in relation to Mountain Ash forests other authors have suggested that moss beds on large logs, self-thinning causing rapid dead fuel

575

accumulation, floristic composition and forest fragmentation at the landscape-scale could contribute to differences in flammability between age-classes (Lindenmayer et al., 2009; Taylor et al., 2014). Further work is required to quantify those other factors in relation to Mountain Ash forests for different fire histories to build a more complete picture of the mechanisms driving flammability in Mountain Ash forests. 5. Conclusion Canopy cover was an important driver of fuel moisture and fuel availability. High canopy cover in the recently burnt high severity site was associated with high fuel moisture, meaning the fuel was rarely available to burn. A sparser canopy in the recently burnt low severity site was associated with low fuel moisture and relatively higher fuel availability. These results suggest that fire severity is an important driver of forest flammability in recently burnt Mountain Ash forest. However, irrespective of the immediate impacts on flammability, all impacts were relatively short-lived. There were no systemic differences in canopy cover, fuel moisture or fuel availability as a function of time since fire or fire severity for the longer unburnt forests (33+ years). Further work is required to quantify other factors that contribute to the overall flammability of Mountain Ash forests. Acknowledgements This research is part of a larger project titled ‘‘Managing bushfire in Tall Mist Forests – fuel hazard and moisture relationships.” The project was managed within the integrated Forest Ecosystem Research program, a research program conducted by the University of Melbourne and funded by the Victorian Government’s Department of Environment, Water, Land and Planning (DEWLP). We gratefully acknowledge the support received from DELWP, Parks Victoria and Melbourne Water to conduct this research. We also acknowledge and thank Geofe Cadiz and Paul Bentley for helping install the weather stations; Dr Matthew Swan, Simon Murphy, Dr April Gloury, Peter Mercouriou and Dr Naomi Davis for conducting tree and vegetation measurements at each weather station; Dr Petter Nyman and Dr Gary Sheridan for technical advice about measuring fuel moisture; and two anonymous reviewers for valuable comments on the manuscript. Appendix A See Fig. A.1.

Fig. A.1. Relationship between manually derived surface, dead elevated and profile fine fuel moisture contents and fuel stick moisture contents. Lines depict 1:1 relationships.

576

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577

References Ashton, D.H., 1976. Development of even-aged stands of Eucalytus regnans F Muell in Central Victoria. Aust. J. Bot. 24, 397–414. Ashton, D.H., 1981. Fire in tall open-forests (wet sclerophyll forests). In: Gill, A.M., Groves, R.H., Noble, I.R. (Eds.), Fire and the Australian Biota. Australian Academy of Science, Canberra, pp. 340–366. Ashton, D.H., 2000. The Big Ash forest, Wallaby Creek, Victoria - changes during one lifetime. Aust. J. Bot. 48, 1–26. Ashton, D.H., Attiwill, P.M., 1994. Tall open forests. In: Groves, R.H. (Ed.), Australian Vegetation. Cambridge University Press, Cambridge, pp. 157–196. Attiwill, P.M., 1994. Ecological disturbance and the conservative management of eucalypt forests in Australia. For. Ecol. Manage. 63, 301–346. Attiwill, P.M., Ryan, M.F., Burrows, N., Cheney, N.P., McCaw, L., Neyland, M., Read, S., 2014. Timber harvesting does not increase fire risk and severity in wet eucalypt forests of Southern Australia. Conserv. Lett. 7, 341–354. Aussenac, G., 2000. Interactions between forest stands and microclimate: ecophysiological aspects and consequences for silviculture. Ann. For. Sci. 57, 287–301. Balch, J.K., Brando, P.M., Nepstad, D.C., Coe, M.T., Silverio, D., Massad, T.J., Davidson, E.A., Lefebvre, P., Oliveira-Santos, C., Rocha, W., Cury, R.T.S., Parsons, A., Carvalho, K.S., 2015. The susceptibility of southeastern Amazon forests to fire: insights from a large-scale burn experiment. Bioscience 65, 898–905. Balch, J.K., Nepstad, D.C., Brando, P.M., Curran, L.M., Portela, O., de Carvalho, O., Lefebvre, P., 2008. Negative fire feedback in a transitional forest of southeastern Amazonia. Glob. Change Biol. 14, 2276–2287. Bassett, O.D., Prior, L.D., Slijkerman, C.M., Jamieson, D., Bowman, D.M.J.S., 2015. Aerial sowing stopped the loss of alpine ash (Eucalytpus delegatensis) forests burnt by three short-interavl fies in the Alpine National Park, Victoria, Australia. For. Ecol. Manage. 342, 39–48. Beniston, M., 2006. Mountain weather and climate: a general overview and a focus on climatic change in the Alps. Hydrobiologia 562, 3–16. Benyon, R.G., Lane, P.N.J., 2013. Ground and satellite-based assessments of wet eucalypt forest survival and regeneration for predicting long-term hydrological responses to a large wildfire. For. Ecol. Manage. 294, 197–207. Blanchi, R., Leonard, J., Haynes, K., Opie, K., James, M., Dimer de Oliveira, F., 2014. Environmental circumstances surrounding bushfire fatalities in Australia 1901– 2011. Environ. Sci. Policy 37, 192–203. Bond, W.J., Keeley, J.E., 2005. Fire as a global ‘herbivore’: the ecology and evolution of flammable ecosystems. Trends Ecol. Evol. 20, 387–394. Bowman, D., Murphy, B.P., Neyland, D.L.J., Williamson, G.J., Prior, L.D., 2014. Abrupt fire regime change may cause landscape-wide loss of mature obligate seeder forests. Glob. Change Biol. 20, 1008–1015. Bradstock, R.A., 2010. A biogeographic model of fire regimes in Australia: current and future implications. Glob. Ecol. Biogeogr. 19, 145–158. Bradstock, R.A., Hammill, K.A., Collins, L., Price, O., 2010. Effects of weather, fuel and terrain on fire severity in topographically diverse landscapes of south-eastern Australia. Landscape Ecol. 25, 607–619. Burrows, N.D., 1999. Fire behaviour in jarrah forest fuels: 2. Field experiments. CALMScience 3, 57–84. Caccamo, G., Chisholm, L.A., Bradstock, R.A., Puotinen, M.L., Pippen, B.G., 2012. Monitoring live fuel moisture content of heathland, shrubland and sclerophyll forest in South-Eastern Australia using MODIS data. Int. J. Wildland Fire 21, 257–269. Chen, J.Q., Saunders, S.C., Crow, T.R., Naiman, R.J., Brosofske, K.D., Mroz, G.D., Brookshire, B.L., Franklin, J.F., 1999. Microclimate in forest ecosystem and landscape ecology - variations in local climate can be used to monitor and compare the effects of different management regimes. Bioscience 49, 288–297. Countryman, C.M., 1977. Radiation Effects on Moisture Variation in Ponderosa Pine Litter. USDA Forest Service, Pacific Southwest Forest and Range Experiment Station, Berkeley, California. Department of Conservation and Environment, 1990. Management Prescriptions for Hardwood Production. Central Gippsland Forest Management Area. Department of Conservation and Environment (Ed.), Melbourne. Department of Environment Land Water and Planning, 2016a. Fire history records of fires primarily on public land. In: Forest, F.A.R., (Ed.). Victorian State Goverment (data.vic.gov.au), Melbourne. Department of Environment Land Water and Planning, 2016b. Modelled old-growth boundaries. In: Land Management Policy (Ed.). Victorian State Goverment (data. vic.gov.au), Melbourne. Estes, B.L., Knapp, E.E., Skinner, C.N., Uzoh, F.C.C., 2012. Seasonal variation in surface fuel moisture between unthinned and thinned mixed conifer forest, northern California, USA. Int. J. Wildland Fire 21, 428–435. Faiella, S.M., Bailey, J.D., 2007. Fluctuations in fuel moisture across restoration treatments in semi-arid ponderosa pine forests of northern Arizona, USA. Int. J. Wildland Fire 16, 119–127. Fetcher, N., Oberbauer, S.F., Strain, B.R., 1985. Vegetation effects on microclimate in lowland tropical forests in Costa-rica. Int. J. Biometeorol. 29, 145–155. Forestry Canada Fire Danger Group, 1992. Development and Structure of the Canadian Forest Fire Behaviour Prediction System. Forestry Canada, Science and Sustainable Development Directorate, Ottawa. Fosberg, M.A., Rothermel, R.C., Andrews, P.L., 1981. Moisture content calculations for 1000-hour timelag fuels. For. Sci. 27, 19–26. Gill, A.M., Zylstra, P., 2005. Flammability of Australian forests. Australian For. 68, 87–93.

Gould, J.S., McCaw, W.L., Chney, N.P., Ellis, P.F., Knight, I.K., Sullivan, A.L., 2007. Project Vesta - Fire in Dry Eucalypt Forest: Fuel Structure, Fuel Dynamics and Fire Behaviour. Ensis-CSIRO and Department of Environment and Conservation, Canberra, ACT. Holdsworth, A.R., Uhl, C., 1997. Fire in Amazonian selectively logged rainforest and the potential for fire reduction. Ecol. Appl. 7, 713–725. Keeley, J.E., Safford, H., Fotheringham, C.J., Franklin, J., Moritz, M., 2009. The 2007 Southern California wildfires: lessons in complexity. J. Forest., 287–296 Koutsias, N., Arianoutsou, M., Kallimanis, A.S., Mallinis, G., Halley, J.M., Dimopoulos, P., 2012. Where did the fires burn in Peloponnisos, Greece the summer of 2007? Evidence for a synergy of fuel and weather. Agric. For. Meteorol. 156, 41–53. Leonard, S.W.J., Bennett, A.F., Clarke, M.F., 2014. Determinants of the occurrence of unburnt forest patches: potential biotic refuges within a large, intense wildfire in South-Eastern Australia. For. Ecol. Manage. 314, 85–93. Lindenmayer, D.B., 2009. Old forest, new perspectives - insights from the Mountain Ash forests of the Central Highlands of Victoria, south-eastern Australia. For. Ecol. Manage. 258, 357–365. Lindenmayer, D.B., Cunningham, R.B., Donnelly, C.F., Franklin, J.F., 2000. Structural features of old-growth Australian montane ash forests. For. Ecol. Manage. 134, 189–204. Lindenmayer, D.B., Hobbs, R.J., Likens, G.E., Krebs, C.J., Banks, S.C., 2011. Newly discovered landscape traps produce regime shifts in wet forests. Proc. Natl. Acad. Sci. U.S.A. 108, 15887–15891. Lindenmayer, D.B., Hunter, M.L., Burton, P.J., Gibbons, P., 2009. Effects of logging on fire regimes in moist forests. Conserv. Lett. 2, 271–277. Long, J.N., Dean, T.J., Roberts, S.D., 2004. Linkages between silviculture and ecology: examination of several important conceptual models. For. Ecol. Manage. 200, 249–261. Long, J.N., Vacchiano, G., 2014. A comprehensive framework of forest stnd propertydensity relationships: perspectives for plant population ecology and forest management. Ann. For. Sci. 71, 325–335. Ma, S.Y., Concilio, A., Oakley, B., North, M., Chen, J.Q., 2010. Spatial variability in microclimate in a mixed-conifer forest before and after thinning and burning treatments. For. Ecol. Manage. 259, 904–915. Matthews, S., 2014. Dead fuel moisture research: 1991–2012. Int. J. Wildland Fire 23, 78–92. Matthews, S., McCaw, W.L., Neal, J.E., Smith, R.H., 2007. Testing a process-based fine fuel moisture model in two forest types. Can. J. For. Res.-Rev. Can. Rech. For. 37, 23–35. McArthur, A.G., 1968. Fire behaviour in eucalypt forests. In: Leaflet No. 107. Department of National Development, Forestry and Timber Bureau, Canberra. McCarthy, M.A., Lindenmayer, D.B., 1998. Multi-aged mountain ash forest, wildlife conservation and timber harvesting. For. Ecol. Manage. 104, 43–56. Meyn, A., White, P.S., Buhk, C., Jentsch, A., 2007. Environmental drivers of large, infrequent wildfires: the emerging conceptual model. Prog. Phys. Geogr. 31, 287–312. Miller, C., Urban, D.L., 2000. Connectivity of forest fuels and surface fire regimes. Landscape Ecol. 15, 145–154. Murphy, B.P., Bradstock, R.A., Boer, M.M., Carter, J., Cary, G.J., Cochrane, M.A., Fensham, R.J., Russell-Smith, J., Williamson, G.J., Bowman, D.M.J.S., 2013. Fire regimes of Australia: a pyrogeographic model system. J. Biogeogr. 40, 1048– 1058. Nyman, P., Metzen, D., Noske, P.J., Lane, P.N.J., Sheridan, G.J., 2015. Quantifying the effects of topographic aspect on water content and temperature in fine surface fuel. Int. J. Wildland Fire 24, 1129–1142. Ough, K., 2001. Regeneration of Wet Forest flora a decade after clear-felling or wildfire - is there a difference? Aust. J. Bot. 49, 645–664. Pausas, J.G., Keeley, J.E., Schwilk, D.W., 2017. Flammability as an ecological and evolutionary driver. J. Ecol. 105, 289–297. R Core Team, 2016. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Ray, D., Nepstad, D., Moutinho, P., 2005. Micrometeorological and canopy controls of fire susceptibility in a forested Amazon landscape. Ecol. Appl. 15, 1664–1678. Rothermel, R., 1972. A mathematical model of predicting fire spread in wildland fuels. In: Research Paper INT-115. USDA Forest Service, Intermountain Forest and Range Experiment Station, Ogden, Utah. Ryan, K.C., 2002. Dynamic interactions between forest structure and fire behavior in boreal ecosystems. Silva. Fenn. 36, 13–39. Slijepcevic, A., Anderson, W.R., Matthews, S., Anderson, D.H., 2015. Evaluating models to predict daily fine fuel moisture content in eucalypt forest. For. Ecol. Manage. 335, 261–269. Sneeuwjagt, R., Peet, G.B., 1985. Forest Fire Behaviour Tables for Western Australia. Western Australian Department of Conservation and Land Management, Perth. Sullivan, A.L., Matthews, S., 2013. Determining landscape fine fuel moisture content of the Kilmore East ’Black Saturday’ wildfire using spatially-extended pointbased models. Environ. Modell. Softw. 40, 98–108. Sullivan, A.L., McCaw, W.L., Cruz, M.G., Matthews, S., Ellis, P.F., 2012. Fuel, fire weather and fire behaviour in Australian ecosystems. In: Bradstock, R.A., Gill, A. M., Williams, R.J. (Eds.), Flammable Australia. Fire Regimes, Biodiversity and Ecosystems in a Changing World. C.S.I.R.O. Publishing, Collingwood, pp. 69–99. Tanskanen, H., Venalainen, A., Puttonen, P., Granstrom, A., 2005. Impact of stand structure on surface fire ignition potential in Picea abies and Pinus sylvestris forests in southern Finland. Can. J. For. Res.-Rev. Can. Rech. For. 35, 410–420. Taylor, C., McCarthy, M.A., Lindenmayer, D.B., 2014. Nonlinear effects of stand age on fire severity. Conserv. Lett. 7, 355–370.

J.G. Cawson et al. / Forest Ecology and Management 400 (2017) 568–577 Thomas, P.B., Watson, P.J., Bradstock, R.A., Penman, T.D., Price, O.F., 2014. Modelling surface fine fuel dynamics across climate gradients in eucalypt forests of southeastern Australia. Ecography 37, 827–837. Tolhurst, K.G., Cheney, N.P., 1999. Synopsis of the Knowledge Used in Prescribed Burning in Victoria. Department of Natural Resources and Environment, Fire Management, East Melbourne. Uhl, C., Kauffman, J.B., 1990. Deforestation. Fire susceptibility, and potential tree responses to fire in the eastern Amazon. Ecology 7. Viney, N.R., 1991. A review of fine fuel moisture modelling. Int. J. Wildland Fire 1, 215–234.

577

von Arx, G., Pannatier, E.G., Thimonier, A., Rebetez, M., 2013. Microclimate in forests with varying leaf area index and soil moisture: potential implications for seedling establishment in a changing climate. J. Ecol. 101, 1201–1213. Wood, S.W., Murphy, B.P., Bowmand, D.M.J.S., 2011. Firescape ecology: how topography determines the contrasting distribution of fire and rain forest in the south-west of the Tasmanian Wilderness World Heritage Area. J. Biogeogr. 2011, 1807–1820. Zou, C.B., Barron-Gafford, G.A., Breshears, D.D., 2007. Effects of topography and woody plant canopy cover on near-ground solar radiation: relevant energy inputs for ecohydrology and hydropredology. Geophys. Res. Lett. 34.