Response of Montane Wet Sclerophyll Forest Understorey Species to Fire: Evidence from High and Low Intensity Fires

2006·3.7 MB·English

by  Monica L Campbell

Response of Montane Wet Sclerophyll Forest Understorey
Species to Fire: Evidence from High and Low Intensity Fires
Monica L. Campbell and Peter J. Clarke
Botany, School ofEnvironmental Sciences andNatural Resources Management, University ofNew England,
NSW
Armidale 2351 ([email protected]).
Campbell, M.L. andClarke, P.J. (2006). Response ofmontanewet sclerophyll forestunderstorey species
totire: evidence fromhighandlow intensityfires. Proceedings oftheLinneanSocietyofNewSouth
Wales 111, 63-73.
OntheNewEnglandTablelandswetsclerophyllforeststypicallyformtheecotonebetweenrainforestand
dry sclerophyll forest. Currentlythere are few dataontheresponse ofwet sclerophyllplant speciesto fire.
Wecomparedthefire-responsetraitsofwoodyunderstoreyandsub-canopyspeciesinwetsclerophyllforest
afterhigh and low intensity fires. The majority ofspecies (>80%) resprouted afterfire andthe prevalence
ofresprouting did not differ with fire intensity. Obligate seeders were rare in these communities (<10%
ofspecies), and similar numbers ofrainforest and sclerophyllous species were killedby fire. Resprouting
from basal stems and root suckering were the most common mechanisms of vegetative regeneration;
however, these traits may have arisen more in response to canopy disturbance than fire regime. We found
that most rainforest taxa resprouted but lacked post-fire seedling recruitment, whereas most resprouting
sclerophyllous taxa recruited from seed after fire. This dichotomy in seedling recruitment could reflect
the productivity and disturbance gradients across the ecotone. We propose that gap-phase recruitment is
favouredtowards therainforestmarginandfire-relatedrecruitmentis moreprevalentatthe eucalyptforest
edge.
Manuscriptreceived 1 May2005, accepted forpublication7 December2005.
KEYWORDS: functional groups, obligate seeding, rainforest, resprouting, seedlingrecruitment, wildfire.
wet sclerophyll forest have been used for hardwood
INTRODUCTION
timber production and cattle grazing. These forests
have consequently been exposed to a regime of
Wetsclerophyllforeststypicallyformaninterface frequent, low-intensity fires that are often associated
between two broad vegetation types, rainforest and with grazing (stimulation ofgreen pick). Ahigh fire-
drysclerophyllforest. Thesetall eucalyptforestshave frequency is thought to reduce diversity of woody
been described as a stage in long-term succession as species in the forest understorey by removing
their structure and composition approaches that of mesophyllous taxa and promoting growth of fire-
rainforest in the prolonged absence of disturbance. tolerant grasses and forbs (Binns 1991; Henderson
Hence, changes in community properties have been and Keith 2002). Currently there are few data on the
closely linked to the passage of fire and length of demographic links between disturbance frequency
fire-fi-ee periods (Ashton and Attiwill 1994). Wet and changes in understorey composition in the wet
sclerophyllforestsarenothighlyflammablemostofthe sclerophyll ecotone.
time, and as time-since-fire increases the probability xhe classificationofplantspecies into fianctional
ofa subsequent fire is reduced as mesophyllous taxa groups based on fire-response fraits can be usefiil in
become more dominant in the standing vegetation preliminary modelling ofhowvegetationwill change
(Unwinl989;Adam 1992; Harringtonand Sandersen with more frequent or less fi-equent fires (Whelan
1994). Generally, one intense crown fire every 100- 1995; Bond and van Wilgen 1996). Baseline data on
200 years is thought to be sufficient to maintain the response of plant populations to crown fire are
the sclerophyllous component and restrict the more being sought to confrast the effects of seed-based
mesophyllous taxa of the wet sclerophyll ecotone recruitment(persistence ofpopulations)withthose of
(Gilbert 1959; Chesterfield et al. 1991). resprouting (persistence of individuals) (Clarke and
On the New England Tablelands many areas of Knox 2002; Pausas et al. 2004; Vesk and Westoby
RESPONSES OF UNDERSTOREY SPECIES TO FIRE
2004). However, generalisations about vegetation grass fire from aneighbouringproperty in October of
change with fire-frequency may be complicated 2001. The fire was oflow to moderate intensity and
by variable responses of species to fire of different resulted in the complete burning of the understorey
intensities (e.g. Ashton and Martin 1996; Morrison andgroundlayer,butwithminimalcanopyscorching.
and Renwick 2000). In addition, shade-tolerant All study sites occurred at altitudes greater than 900
m
rainforest species are not expectedto have fire-driven on metasediment-derived soils. Prior to the fires,
recruitment, and their regeneration syndromes are areas of wet sclerophyll forests within each park
more likely to be linked to small-scale disturbances were surveyed for full floristic composition using 20
m
such as tree fall that create light gaps. X 20 quadrats (21 vegetation survey sites in total:
In spring 2002, areas ofwet sclerophyll forest in WPNP, 12 sites; MGNP, 9 sites). Vegetation in each
Washpool National Park were burntby fire following quadrat was described in terms of the growth form,
the severe drought that affected most of eastern height and cover of dominant species in the ground
Australia. We took advantage of this one in fifty storey, understorey and canopy strata. All species
year event to record the response ofwet sclerophyll were recorded in each quadrat and their abundance
understorey species to crown fire. We focused on the estimated using Braun-Blanquet cover-abundance
shrubandsub-canopytaxa,asthedominantoverstorey scale. For each survey quadrat geographical
eucalypts all resprout after fire and their dynamics position, altitude, slope, local soil characteristics,
have been documented elsewhere (e.g. Ashton and physiography, evidence offire and otherdisturbances
Attiwill 1994; Florence 1996). To test the generality were recorded.
ofresponses to fire we compared datafrom the crown
fire at Washpool National Park with a lower intensity Responses ofadult plants to fire
bum at Mummel Gulf National Park the previous Responses of woody plant species to fire were
year. We addressed the following questions: 1) Do recorded at four sites within burnt forest at WPNP
fire-responsefraitsvarywithfire intensity?2) Dofire- and MGNP. At each ofthese sites two transects (20
response traits varybetween sympatric rainforest and X 2 m) were placed along each ofthree topographical
sclerophyllous species? and 3) Are there correlations positions - ridge, slope and gully. Topographical
between environmental variables, fire-response traits position accounted for differences in vegetation
and other life-history traits? composition prior to fire and differences in fire
intensity. Post-fire responses of adult plants were
recorded along each transect in four categories:
METHODS
1) killed by fire, 2) resprouting via root suckers, 3)
resproutingviabasalstembuds, and4)resproutingvia
Study areas stembuds. The presence ofpost-fire seedlingrecraits
The New England wet sclerophyll forests was also recorded. In addition to these observations,
are restricted to areas of high rainfall along the post-fire responses ofother woody species outside of
eastern edge of the escarpment. These forest are transects (16 species) were recorded to gain a more
characterised by a eucalypt-dominated overstorey, comprehensive overviewoffire-responsetraits ofwet
typically exceeding 30 m in height, and a well- sclerophyll forest understorey species.
developed, layered understorey ofmesomorphic and
sclerophyllous growth forms (Specht 1970; Ashton Analyses offire response traits
and Attiwill 1994). The area selected for study at Plant species were allocated to one offive fire-
Washpool National Park (hereafter WPNP) was in response syndromes based on categories defined by
the recent western additions that were acquired by Gill and Bradstock (1992). Note that because ofthe
the New South Wales National Parks and Wildlife low frequency of occurrence, species killed by fire
Service (NPWS) in 1998 and had not been burnt for were classified into one group regardless of seed
at least 50 years. The fires that occurred inNovember banktype. An additional group was also formed with
2002werehigh-intensityfiresthatburntallvegetation the combination of regeneration by root suckering
strata in the sclerophyll forests and the understorey (Category IV) and resprouting by basal stem buds
of the warm temperate rainforest. Mummel Gulf (Category V). Fire-response traits ofwoody species
National Park (hereafter MGNP) lies approximately were compared between sites of different fire
60 km east of Walcha and was also gazetted as intensities, i.e. high-intensity fire (WPNP) and low-
National Park in the late 1990s. The fire at MGNP intensity fire (MGNP), and between topographic
WPNP
was a back-bum, initiated by the NPWS to contain a positions (ridge, slope and gully) within and
MGNP. The relative frequency offire-response traits
64 Proc. Linn. Soc. N.S.W., 127, 2006
CAMPBELLAND CLARKE
M.L.
P.J.
Table 1, Summary table for life-history attributes foliage types; sclerophyllous, coriaceous
of 61 woody taxa occurring in wet sclerophyll for- and mesophyllous. These categories were
est in Washpool and Mummel Gulf National Parks on chosen as preliminary representatives
the New England Tablelands. Note eucalypts not in- of the different life-history syndromes
cluded. Figures are the No. of species in each category. of plant species in the wet sclerophyll
ecotone. Rainforest taxa generally fell
into the mesophyllous and coriaceous leaf
WPNP MGNP
Attribute types, while most sclerophyll forest taxa
Growth form were classified as sclerophyllous. Note that
because of low frequencies of occurrence,
Shrub (<3 m) 17 19
taxa with coriaceous leaves were grouped
SmallTree (3-10 m) 13 9 with the mesophyllous taxa for the analysis.
Tree (>10m) 12 6 The frequencies of persistence syndromes
(as described above) were tested between
foliagegroupswithaG-testforindependence
Leaftype (Sokal and Rolf 1981).
Sclerophyllous 15 15
Life-history traits and environmental
Coriaceous 6 7
variables
Mesophyllous 21 12 Toexaminetherelationshipbetween
plant traits and enviroimiental variables,
constrained ordinations were derived from
Dispersal syndrome
the full floristic data set. Patterns of life-
Vertebrate 21 21 history traits and envirormiental correlates
Invertebrate 9 4 were testedwith Canonical Correspondence
CANOCO™
Analysis (CCA)using v4.5 (ter
Wind 4 5 A
BraakandSmilauer 1992). life-historytrait
Passive 8 4 data set was constructed using observations
in the field and from other sources. Growth
form, leaf type, dispersal syndromes and
Seedbanktype
seed bank classes are shown in Table 1.
Soil 31 26 Plant persistence syndromes were also
Canopy 3 3 included in the analysis. An environmental
matrix was constructed using the variables
Dispersed 8 5
recorded in each vegetation survey
quadrat and climatic parameters extracted
Species richness 42 34 from BIOCLIM™ (Busby 1991). For the
purposes of analysis, Australian Map Grid
Eastings andAustralianMapGridNorthings
was compared with a G-test for independence (Sokal were included as covariables. The significance of
and Rolf 1981). Fire-response traits of adult plants envirormiental variables was determined using 499
after crown and understorey fires were reclassified Monte Carlo Permutations and the forward selection
intooneofthefourplantpersistencesyndromesbased option in CANOCO™ v4.5 (ter Braak and Smilauer
on the hierarchical persistence scheme of Pausas et 1992).
al. (2004). The relative fi-equencies of persistence
syndromes after crown and understorey fire were
compared with a G-test for independence (Sokal and RESULTS
Rolf 1981).
Fire-response traits, persistence syndromes and
Analyses ofpersistence syndromes and foliage fire intensity
types Data on the fire-response of 49 woody taxa
To test for differences in persistence syndromes were collected from transects at WPNP and MGNP.
betweenrainforestandsclerophyllforesttaxagrowing Resprouting was the dominant response to fire (>
in the same habitat, species were divided into three 80% oftaxa. Table 2) and there was no difference in
Proc. Linn. Soc. N.S.W., 127, 2006 65
RESPONSES OF UNDERSTOREY SPECIES TO FIRE
Table 2. Summary ofcontingency-table analysis for fire-response traits ofwoody understorey species
following high intensity and low intensity fires in National Parks on the New England Tablelands. Fig-
ures are the No. ofspecies in each category.
High intensity fire (WPNP) Low intensity Fire (MGNP)
Ridge Slope Gully Total Ridge Slope Gully Total
^^W
Fire Response (II-VI)
II. Killed, soil-stored seed bank 5 5 4 6 4 1 1 2
IV. Resprouts via root suckers 3 4 4 3 1 2 4 3
V. Resprouts via basal stem 13 14 12 16 10 14 13 18
VI. Resprouts via stem bud bank 1 1 1 1 1 1 1 1
& V
Both IV 5 4 5 8 3 4 4 3
Fire Response
Killed (> 70% killed) 5 5 5 6 3 1 1 2
Resprout (< 30% killed) 21 23 20 27 15 20 21 22
70%
Variable (30 - killed) 1 1 1 1 1 1 3
the frequency of fire-response traits between crown the landscape, although a greater number of species
fire (WPNP) and understorey fire (MGNP) (G^ = killed by fire were recorded at the high-intensity fire
2.68, P > 0.05). The fi^equency of resprouting traits sites (WPNP, 17%) than at the low-intensityfire sites
was also consistent across topographic gradients (MGNP, 7%) (Table 2).
within parks (WPNP, G^ = 0.87, P > 0.05; MGNP G^ Additional observations recorded outside
= 4.19, P > 0.05). Resprouting firom basal stems was transects were included in the persistence syndrome
the most common fire-response, followed by species data set, with persistence traits of 54 woody species
regenerating from both basal stems and root-suckers included in the analysis. Resprouting was the most
and then those regenerating from root-suckers alone common persistence syndrome, although resprouting
(Table 2). Species killedby fire were less common in without post-fire seedling recruitment was more
frequentthanresproutingwith
Table 3. Summary table of contingency analysis for persistence syn-
post-fire seedling recruitment
dromes ofwoody understorey species after crown and understorey fires (Table 3). Species killed by
and between leaf-type classes. Persistence syndromes defined by Pau- fire were low in frequency
sas et al. (2004), presence or absence of seedlings refers to post-fire re- at both sites. There was no
cruitment only. Species killed by fire and lacking post-fire recruitment
significant difference in the
rely on the dispersal of propagules into a burnt area for recruitment.
relative frequencies of any
ofthe persistence syndromes
Low intensityHigh intensity Sclerophyllous Mesophyllous/ between high-intensity fire
(WPNP) (MGNP) Coriaceous sites and low-intensity fire
..iv»'^-i . P>
1 sites (G2= 2.84, 0.05).
Resprouters
+ seedlings 9 10 12 1 Persistence syndromes and
- seedlings 21 15 2 29 leaftypes
Sclerophyllous (dry forest
taxa) species had a higher
Killed frequency of resprouting
+ seedlings 6 2 3 4 species with post-fire
seedling recruitment than
- seedlings 1 2 1 2
mesophyllous and coriaceous
66 Proc. Linn. Soc. N.S.W., 127, 2006
.
M.L. CAMPBELLAND PJ. CLARKE
Figure 1. Biplot diagram of CCA ordination for environmental variables and
life history traits of 61 woody taxa in wet sclerophyll forest. D= life history traits;
solid arrows = significant environmental variables (499 Monte Carlo Permu-
tations, P< 0.05). RS-, RS+, K-, K+ = persistence syndromes defined by Pau-
sas et al. (2004); passive, wind, vertebrate, invertebrate = dispersal syndromes;
Spp Rich = species richness; AMMI = annual mean moisture index; MICV =
moisture index coefficient of variation. Plot axes are 1 by 1 units of ordination.
taxa {&= 22.66, P < 0.0001) (Table 3). Conversely, more common at WPNP (Table 1). There were
rainforest taxa (mesophyllous and coriaceous plants) more sclerophyllous and mesophyllous taxa than
had significantly greater frequency of species that coriaceous at both sites and most species had
resprouted and lacked post-fire seedling recruits vertebrate-dispersedpropagules (Table 1). Soil-stored
(Table 3). There was no difference in the frequency seed banks were the predominant type with less than
of species killed by fire with or without post-fire 20% of species having a canopy-held seed banks or
seedling recruitment between leaftype groups (Table relying on dispersal of seed for regeneration (Table
3). 1). More detailedinformation onthe life-historytraits
ofindividual species is given inAppendix 1
Plant traits and environmental variables Life-history traits of woody taxa in the wet
Shrubs were the most common woody growth sclerophyll understorey were significantly correlated
forms at both sites, but trees and small trees were with canopy cover, slope and understorey cover
Proc. Linn. Soc. N.S.W., 127, 2006 67
RESPONSES OF UNDERSTOP^Y SPECIES TO FIRE
(Fig. 1). Mesophyllous and tree attributes correlated the low frequency of obligate seeding species and
with increasing canopy cover, as did resprouting post-fire recruitment in these communities. We also
species that lack post-fire seedling recruitment (Fig. recorded high levels of root-suckering, and roughly
1). Species killed by fire and with post-fire seedling one quarterofthe specieswere capable ofresprouting
recruitment irom in situ seed banks were negatively from the basal stem and roots after fire. Resprouting
associatedwithincreasingcanopycover,butpositively from basal stem tissue suggests that moderate fire-
correlatedwith increasing ground cover. Resprouting frequencies have been a selective force (Bellingham
species with post-fire seedling recruitment were and Sparrow 2000), but resprouting may also confer
weakly associated with increasing understorey cover, an advantage by enabling individuals to survive and
but correlated closely with moisture predictability regenerate aftermechanical damage inflicted by tree-
and temperature (Fig. 1). Species with canopy-stored and limb-fall (Ashton 2000; Paciorek et al. 2000;
seed banks and wind-dispersed seeds were positively Kanno et al. 2001). Similarly, root-suckering is an
correlatedwith landscape slope (Fig. 1). Species with effective means of invading unoccupied space after
passive dispersal, those lacking in situ seed banks, disturbance events such as tree-fall (Stocker 1981;
and those killed by fire with seeds dispersed into the Kammesheidt 1999; Bellingham and Sparrow 2000).
post-fire environment, were all negatively correlated
with increasing understorey cover (Fig. 1). Do fire-response traits vary between rainforest
and sclerophyllous species?
There was a clear dichotomy between rainforest
DISCUSSION (mesophyllous and coriaceous) and sclerophyllous
taxa in relation to post-fire seedling recruitment. The
majority of sclerophyllous taxa that resprouted also
Do fire response traits vary with fire intensity?
We found that the majority (>80%) of woody had post-fire seedling recruitment, whereas most
rainforesttaxaresproutedbutlackedpost-fireseedling
understoreyspeciesinmontanewetforestsresprouted
recruitment. This may be explained by resource
after fire irrespective of fire intensity. Whilst fire
gradients across thewet sclerophyll ecotone affecting
intensity was not replicated across sites our general
species composition. At the rainforest interface the
observations in wet forests in the region support this
quantity and quality oflight reaching the forest floor
finding. This contrasts with observations that high-
is much lower than at the eucalypt forest edge and
intensity fires limited vegetative regeneration in
these conditions are generally unfavourable for the
Victorian wet sclerophyll forests, where rootstocks
recruitment ofshade-intoleranttaxa (Turton andDuff
and bud banks did not survive high temperatures.
1992). The prevalence ofmesophyllous species at the
Conversely,afterlower-intensityfire,thepredominant
rainforest interface reduces theprobability offire and
mechanism of regeneration was vegetative with a
species with gap-phase regeneration dominate the
high density ofroot-suckers andresprouting adults in
community (Unwin 1989; Adam 1992; Harrington
the post-fire environment (Ashton and Martin 1996).
and Sandersen 1994). The general absence of post-
In our study, less than 10% of species demonstrated
fire seedling recruitment in rainforest taxa is likely
a variable response to fire, indicating that the
to reflectthat recruitment syndromes in rainforest are
dichotomous classification of species into obligate
linkedto canopydisturbanceratherthan^reperse. In
seeders (< 30% resprout) and resprouters (> 70%
contrast, post-fire seedling recruitment was common
resprout) is auseful generalisation for these systems.
in sclerophyllous taxa: these species respond to
The proportion of resprouting species recorded
broad-scale disturbance in order to regenerate and
here is high in comparison to wet sclerophyll
gap-phase recruitment is rare (Melick 1990).
forests in Victoria (30%) and southwest Western
The dichotomy in seedling recruitment
Australia (24%), but comparable to data for coastal
wet sclerophyll forest in northern New South Wales syndromes was reflected in the CCA, with species
that lack post-fire recruitment closely associated
(60%) (Ashton 1981). Similarly high proportions of
with increasing canopy cover. In contrast, species
resproutingspecieshavebeenreportedforotherhighly
with post-fire seedling recruitment, regardless ofthe
competitive systems such as wet heaths and grassy
woodlands on the New England Tablelands (Clarke adultplantresponsetofire, werepositivelyassociated
with more open habitats as indicated by increasing
and Knox 2002; Clarke et al. 2005). Resprouting
groundandunderstorey cover.Atthe landscape scale,
may be favoured in productive habitats as vegetative
productivity gradients have been linked to ratios
recruitsandregenerationarecompetitivelysuperiorto
of obligate seeders to resprouters across habitats
seedlings (Clarke et al. 2005). This may also explain
(e.g. Clarke et al. 2005). However, within the wet
68 Proc. Linn. Soc. N.S.W., 127, 2006
CAMPBELLAND CLARKE
M.L.
P.J.
sclerophyll ecotone,theproductivitygradientappears microcorys—E. salignawet sclerophyll forest in
to be driving the prevalence of post-fire seedhng responseto logging. MSc Thesis, University ofNew
recruitment more than adultfire responses. England,Armidale.
Paradoxically, mostrainforestspecies vigorously Bond, W.J. andvanWilgen, B.J.R (1996). 'Fire and
plants'. (Chapman and Hall, London).
resprouted and similar numbers of rainforest and
Busby, J.R. (1991). Bioclim-abioclimate analysis and
sclerophyll species were killed by fire. Other
prediction system. PlantProtection Quarterly 6, 8-9.
studies have reported rainforest species coppicing Chesterfield, E.A., Taylor, S.J. andMolnar, CD. (1991).
or resprouting after fire (Stocker 1981; Chesterfield Recovery afterwildfire: warm temperate rainforest
et al. 1991; WiUiams 2000). Hence the notion of a atJones Creek, East Gippsland,Victoria.Australian
split 'fire-intolerant' vs. 'fire-tolerant' flora does ForestrySA, 157-173.
not appear to be explained simply by differences in Clarke, PJ. andKnox, K.J.E. (2002). Post-fire response
resprouting ability. Recently, however, Fensham et ofshrubs inthe tablelands ofeasternAustralia:
do existingmodels explainhabitat differences?
al. (2003) demonstrated that recurrent fires caused
AustralianJournalofBotany50, 53-62.
increased mortality in tree species from monsoon
Clarke, RJ., Knox, K.J.E., Wills, K.E. andCampbell, M.
rainforest compared to surrounding savannah,
(2005). Landscape patterns ofwoodyplantresponse
suggesting fundamental differences in sprouting to crownfire: disturbance andproductivityinfluence
ability. Remaining unresolved is the question of sproutingability. JournalofEcology93, 544-555.
whether quintessential sclerophyllous species are Fensham, R. J., Fairfax, R.J., Butler, D.W. andBowman,
more 'fire-tolerant' than their mesophytic cousins in D.J.M.S. (2003). Effects offire and drought in a
the same genus or family, and what the mechanisms tropical eucalypt savanna colonizedbyrainforests.
forthis tolerance are. JournalofBiogeography30, 1405-1414.
Florence, R.G. (1996). 'Ecologyand silviculture of
eucalyptforests'. (CSIRO Publishing, Collingwood).
ACKNOWLEDGEMENTS Gilbert, J.M (1959). Forest successioninthe Florentine
Valley, Tasmania. ProceedingsoftheRoyalSocietyof
Tasmania 93, 129-151.
We thank Richard Willis and Shanti Virgona for Gill,A.M. and Bradstock, R.A. (1992).Anafional register
assistance in the field. Financial support was provided to forthe fireresponse ofplantspecies. Cunninghamia
MLC by an Australian Post-graduate Award (Industry), 2, 653-660.
New South Wales National Parks and Wildlife Service, Harden, G. J.E. (1990). 'Flora ofNew SouthWalesVol.
University ofNew England and NCW Beadle scholarship. 1'. (NSWUniversityPress, Kensington).
Lachlan Copeland kindly assisted with the identification Harden, G.J.E. (1991). 'FloraofNew SouthWalesVol. 2'.
andnomenclature ofplanttaxa. (NSWUniversityPress, Kensington).
Harden, G.J.E. (1992). 'Flora ofNew SouthWalesVol. 3'.
(NSWUniversityPress, Kensington).
,
REFERENCES Harden, G.J.E. (1993). 'FloraofNew SouthWalesVol. 4'.
(NSWUniversityPress, Kensington).
Harrington, G.N. and Sandersen, K.D. (1994). Recent
Adam, P. (1992). 'Australianrainforests'. (Clarendon
contraction ofwet sclerophyll forest inthewet
Press, Oxford).
tropics ofQueensland due to invasionbyrainforest.
Ashton, D.H. (1981). Fire intall open-forests (wet
Pacific ConservationBiology 1, 319-327.
sclerophyll forests). In 'Fire andtheAustralianbiota'
Henderson, M.K. andKeith, D.A. (2002). Correlation
(EdsA.M. Gill, R.H. Groves andI.R. Noble) pp. 339-
ofburning andgrazing indicators with composition
366. (AustralianAcademy ofScience, Canberra).
ofwoodyunderstoreyfloraofdells in atemperate
Ashton, D.H. (2000). The BigAsh forest, Wallaby Creek,
Victoria-changes during one lifetime.Australian eucalyptforest.AustralEcology27, 121-131.
Kammesheidt, L. (1999). Forestrecoverybyroot
JournalofBotany4S, 1-26.
suckers and above-ground sprouts afterslash-and-
Ashton, D.H. andAttiwill, RM. (1994). Tall open-forests.
bum agriculture, fire and logging in Paraguay and
In 'AustraUanVegetation' (Ed. R.H. Groves) pp. 157-
Venezuela.JournalofTropicalEcology 15, 143-157.
196. (CambridgeUniversityPress, Cambridge).
Kanno, H., Hara, M., Hirabuki,Y, Takehara,A. and Seiwa,
Ashton, D.H. andMartin, D.G. (1996). Regeneration in a
K. (2001). Populationdjoiamics offourunderstorey
pole-stage forestofEucalyptus regnans subjectedto
shrub species during a 7-yrperiod inprimarybeech
differentfire intensities.AustralianJournalofBotany
forest. JournalofVegetation Science 12, 391-400.
44, 393-410.
Melick, D.R. (1990). Ecology ofrainforest and sclerophyll
Bellingham, P.J. and Sparrow,A.D. (2000). Resprouting
communities inthe Mitchell RiverNational Park,
as a life history strategy in woodyplant communities.
Gippsland, Victoria. Proceedings oftheRoyalSociety
Oiito5 89, 409-416.
ofVictoria 102, 71-S7.
Binns, D.L. (1991). Vegetation dynamics oiEucalyptus
Proc. Linn. Soc. N.S.W., 127, 2006 69
RESPONSES OF UNDERSTOREY SPECIES TO FIRE
Morrison, D.A. andRenwick, J.A. (2000). Effects of
variationinfire intensityonregeneration ofco-
occurring species ofsmall trees inthe Sydneyregion.
AustralianJournalofBotany48, 71-79.
Paciorek, C.J., Condit, R., Hubbell, S.R andFoster, R.B.
(2000). The demographics ofresprouting intree and
shrub species ofamoisttropical forest.Journalof
Ecology88, 165-111.
Pausas, J.G., Bradstock, R.A., Keith, D.A. andKeeley,
J.E. (2004). Plant functionaltraits inrelationto fire in
crown-fire ecosystems. Ecology85, 1085-1100.
Sokal, R.RandRolf, F.J. (1981). 'Biometry'. (Freeman
andCompany, SanFrancisco).
Specht, R.L. (1970). Vegetation. In 'TheAustralian
environment' (Ed. G.W. Leeper)pp. 44-67. (CSIRO B .- S
andMelbourne UniversityPress, Melbourne). ? ^ I
K ^ H 5
Stocker, G.C. (1981). RegenerationofanorthQueensland E "^ ^ ««
^
rainforestfollowingfelling andburning.Biotropica <S, Si ^
13, 86-92. '^ 1 "S 3
terBraak, C.J.F. and Smilauer, R (1992). 'CANOCO "S, J s "I
version4.5 reference manual & users guideto 2 ^ "^ S
CANOCO forWindows'. (MicrocomputerPower, "3 ^ '.^ |
Ithaca, NY). ^ s § Z
Turton, S.M. andDuff, G.A. (1992). Lightenvironments I — S S
andfloristic composition across open forest-rainforest « S *« "S
boundary innortheastern Queensland.Australian 5 S S" 5
JournalofEcology \1,A\5A23. o 2 « ^
Unwin, G.L. (1989). Structure and composition of ^ « y «
the abruptrainforestboundary inthe Herberton vo § -3 "O
Highland, NorthQueensland.AustralianJournalof *© •" ?
"o JJ^
Botany37, 413-42%. ^ g, *^ g^^g;
Vesk,P.AandWestoby,M. (2004). Globalpatterns of 1 *
a^ a«\
sprouting ability: can allplant species be dividedinto '
Sprouters andnon-sprouters?JournalofEcology92, 1^- « ^ ^-
310-320. .a ««. rf«^ (« ov
Whelan,R.J. (1995). 'Ecology offire'. (Cambridge ^ _ _ -
UniversityPress, Cambridge). "(^jN aflj. ^f* *fM i' T•
Williams, P.R. (2000). Fire-stimulatedrainforest seedling i-h '^ s' * o^n
recruitment andvegetativeregeneration in adensely .S o '^^ S
^
grassedwetsclerophyll forestofnorth-eastern "g « « §
^
Australia.AustralianJournal ofBotany48, 651-658. O. -S cc "2
3'^^^ * 5
-< s * PC
70 Proc. Linn. Soc. N.S.W., 127, 2006
M.L. CAMPBELLAND CLARKE
P.J.
«
zOh zdn Z&, zCL, Zeu Ozh 0Z. 1 1
^eu ^Pu ^CL| ^a. 5c^u Oh O^h O^h 5CL:.
z z oZT oZT aZ. oZT CzL. cZC 0z- CZlT aZ, Ozh oZ. CzU Ozh oZ- CZU Ozh' oz- Ozh oz- Ozh OZh' Ozh' Ozh
A3.= O^^ o2 a2 a2 os as os 05.: aS os Oh o Oh Oh a o O^h sC3 ^Oh so o^. 5O^h as sa ^CU
i •= +
> > > > >
S S > > > K > > =
> > > > >
IX
^ a a
CcCQB ^o^ ^o^ _o^ ^o^ ^o^ ouS2 ^o^ o, ^o^ O>> >> 1) 'S 3 '0 '0 'S '5 '0 '0 '0 'S 'S
t>0 (Zl C/3 c« Ol .2 c/o (Zl CO g c3 !Z1 1/5 !/3 t/3 tzi t/3 .1 IZl »2
(U
o Xcuctuu>cIdH .euu>uCtcCOH JcUu>cUa2dh Xcr>UcIth Xc"u>icI5M X>JcuCh!3S.> X>JcU1U>U3i >jXceUwu5s C_«5MCU>^/^5 T§C3 -^ca X>^3Cr1;)Dt). CJUO>US] C<c0>Uua5 X>Jcu«Ua5U X>Jcu«biI<> J>J^tc2U;. X>^"tubuciw^d X>.tt<L>!dh> "OrUC>rhO> >X^"cUtIw^H *Ou>C«h8 'C>CwUB X>JtrCU)t^ X>^"duuuSS- X>^at<tu3wL:>
x>L:h. x§>:. x>:^ x>:> Js>3. sa. xI>:-h, JS05, ss J1>=. x§U>:i. 20, js3 SDg>1. g8CDO xQO;. xao:, O xBOC:O. '.uC38-COH 'U.C3§CC—OJ xsww&: '.u3gC2 *.uug3c2 2aCSf>l< x2tg>:/3>.
tzi on 00 00 00
O
o -- -- r^ r^ X5I Xx023:0 X0I0 X2) H Ecd MEA EA AS EcQ AE r^ =—^ f<iu: ^2
o C/D C/l 05 t/3 (Zl
•1
1
|
ua. <C3C/O3 1.<1C2O '2.<E52 1<ECXO *1]..<322 <ECCiOO Xta0OC(3oOU0S .x•s1cE^>a« .1tuf2C22ocO X.1f.c3C)22e«3 *CO 'CE23SO _31'_"TCCu2Co>303OAO .1"2CE2aO "U>EE3cco •t*TCC§CEC1acc3XOOOI cU'C8>C1SSOO^ 1iS0CC80OOS0. '•1Cc5cOSb -|«1XQCgCal>OO 15tCC2C/OOO5 *aQC0CauECc&OOO0o exQC12&g:O> "m2CuC3eC8C3OOOO- 1•.•E2c§25 .JWn32.
C/5 <; <: CQ CQ DQ PL)
CO <u
CO CuO
E xtCCCu>uJOO XPutCCs)u)OO XpCutCCl)uOjOOU XtCuC<C5uOODO XCC(CC)UOUOO uCCJO PuCDJO. c 3 C2OAUJ xQC0L3C:QhO. 1UcC<3Ou '.Ue3yS2 h2(Cr3<JOtUU ^B23CiO h2a3CCJO>O J<32CCDOO XP)U •ca pj Ji3C:Oi P03U.
Proc. Linn. Soc. N.S.W., 127, 2006 71
RESPONSES OF UNDERSTOREY SPECIES TO FIRE
&Z< CZ^ ag,
o a a a a o o a o o o
ph
S S ^
« £
o
Oh
> > > >
=a > > '^ > > •^ > ^ >
> > t >
a
-0ea3a Co/3 Uoc to/l 0o0 Co/2 UfocSt U5coc o 1O/3 (CV(f/UhU3t TOUC;f3-Ait T<Mt1f3LH))t 0o0 TUMt1f-3)it -<cl<fo-u«uit 0o0 0o0 0o0 0o0 Co/5 oon 1o/3 (o/5
Q Q
•oa
o sued '-c$a J2a2> aca ^c2a> xSBU> ^TC3 > ^^1o ^^2a> •^CT3 ^TC3 •^Ba X^§a)> '-a$a > X1aI> X^Bu3 Xisu) X^1u) Ss> 1MW X^Bu3 Xu!
ft ^ > ^ a. ^ Ph (X Oih^ >
g ;^ ;^ ;^ :^ ;^ :^ :^ ;^
fO>t1 ofot xfou:t fO(oAt xjr>:j. _x&>B:-, J<fo>SDt> Jafo>S>t, x: •UoO3c •UoO3c .Ucoog2 cU0oo3^ sgou. x1f?:-t. gso. xf2>:-t. Jfa?=-t. x7f2>:3t. x1fu>:t> jsfos3t xf2>:t. JSfo3t J1f>3t.
00 oo 00 00 00 00 00 V3 00 00 OO
o
I "3 "3 F^^-r 1.^^ 1-J-^ X23 I
a2 OO VI aB aB xo:o XC/Sl cEa
00 !/3
s
^
a> O .2 CO
0f0t .".aooc132d Mc.fucucccfaaaaai fo^ogcaf>uai. ^sC2oaCc3Oap 1lCoMOfOo3(-UJt _JuCcooCfoou32itO .cEo2a X"cosofO5t 2foot o a>o ofocfatt xO3cE:a •Oc o 'COOCO§oobSBA« sOooChA BcOOCU<ELLiiA> PooCULsiA ccEaa "OuacEuIa 6SofCoAt X_"-OaOh3oo
o c(aU «<3U cua
s cEa u a1^ Ca c(CaUJ dcuu^a uQ« Uu« o oft
i Xcc)aa f<2t1> 1 tin wft Wocfat S">t^> O^cia XWf33t o;-l -^s o)-• o(-1 ou o^ o Xc)a
72 Proc. Linn. Soc. N.S.W., 127, 2006