Intense Fires Promote Uncommon Post-fire Ephemerals In Currawang 'Acacia Doratoxylon' Dry Scrubs of Little River Gorge, East Gippsland

2007·3.3 MB·English

)
Research Report
Intense fires promote uncommon post-fire ephemerals in
CurrawangAcacia doratoxylon dry scrubs of
Little River Gorge, East Gippsland
Suzanne M Prober', Kevin RThiele’ and M Bramwell"
'CSIROSustainableEcosystems,PrivateBag5,Wembley,WesternAustralia6913
'DepartmentofEnvironmentandConservation,LMB 104,BentleyDeliveryCentre,WesternAustralia6983
'DepartmentofSustainabilityandEnvironment,574MainStreet,Bairnsdale,Victoria3875
Abstract
Acacia scrubs are an unusual vegetation community in the Upper Snowy River region ofVictoria
and New South Wales, occurring as sharply-delineated patches amongst eucalypt forest on steep,
rocky, north towestfacing slopes. Theycan be dominatedbyarange ofAcacia species, butunder-
storeys ofAcacia scrubs are generally reported as sparse and species poor, and little is known of
their ecology. Extensive wildfires in 2003 provided an opportunity to compare the effects ofmild
and intense fireson plantspecies composition and vigourinAcaciascrubcommunities ofthe Little
River Gorge in eastern Victoria. Acacia scrubs atthis site were dominated byAcaciadoratoxylon,
notpreviouslydescribedasadominantinAcaciascrubsinVictoria.Consistentwithearlierobserva-
tions, understoreys in unburnt scrubs were usually sparse. Light burns led to minimal change in
understoreyabundance,andsporadicestablishmentofarangeofspeciesthatwere uncommon inthe
understorey ofunburnt sites. By contrast, burns that were intense enough to kill the Acacia over-
storey stimulated dense establishmentand profuse growth ofspeciesthatwere rare or absent in the
understoreyofunburntplots, leadingtoadramatic increase in nativeshrubrichnessandunderstorey
abundance. Processes leading to these understorey changes are likely to involve both effects offire
on seed dormancy, and release from intense competition for resources after overstorey death.
Infrequent, intensefireshavepreviouslybeenrecognisedas importantforstimulatingrecruitmentof
overstorey species and thus the perpetuation ofAcacia scrubs in far south-eastern Australia. This
study indicated that a diverse range ofunderstorey species are also dependent on intense fires in
these communities, where they appear to behave as post-fire ephemerals. Further, many ofthese
species are listed as rare or vulnerable in Victoria, and thus the appropriate fire management of
Acaciascrubsisofsignificantconservation interest. (The VictorianNaturalist124(6),2007,320-331
Introduction
Dry sclerophyll forests or ‘scrubs’ dominat- kettlewelliae and A. silvestris') have been
ed by Acacia species form an anomalous recorded as dominants in different scrub
vegetation community in the Snowy River patches (Costin 1954, Clayton-Greene and
region of eastern Victoria. They oceur as Wimbush 1988, Keith and Bedward 1999).
sharply-delineated patches on steep, rocky In the Snowy River region, Clayton-Greene
northandwestfacingslopeswithinamatrix andWimbush (1988)foundAcaciasilvestris
ofeucalypt-dominated vegetation. Together was dominant in most stands. The under-
with related scrubs ofsouth-eastern NSW, storey ofAcacia scrubs is usually reported
these scrubs form the only class ofvegeta- as sparse,with scattered shrubs andminimal
tion east ofthe Great Dividing Range in ground cover, but a dense shrub layer, par-
which Acacia is the dominant tree (Keith ticularly ofPhilotheca trachyphylla (some-
2004). They are not closely related to any times without the Acacia overstorey) can
other classes ofvegetation in the region, also occur (Forbes et al. 1982, Clayton-
although they share some floristic similari- Greene and Wimbush 1988, Keith 2004).
ties with rocky outcrops in the adjacent Keith (2004) describes the community as
sclerophyll forests (Keith 2004). In eastern typically species-poor, with only about a
Victoria, they are known as rocky outcrop dozen speciesper0.1 hectare.
scrubs(EVC27, Woodgateetal. 1994; pre- Little is known about the ecology ofdry
viously rocky-outcrop open scrubland com- Acacia scrubs (Clayton-Greene and
munityofForbesetal. 1982). Wimbush 1988). The even-aged nature of
Various Acacia species (A. binervia, A. matureAcacia stands indicates that intense
blayana, A. doratoxylon, A.falciformis, A. fires are important for their regeneration.
320 The Victorian Naturalist
Research Report
In these conditions, mature trees are killed fuel loads and weather conditions, leading
and gennination andgrowth ofseedlings is to a mosaic of unburnt, lightly burnt and
stimulated (Clayton-Greene and Wimbush intensely burntareas acrossthegorge.
1988). However, there is little documenta-
Methods
tion of effects of fire on other species in
Sampling
these communities, and patterns in species
composition relating to other environmen- Floristic sampling focused on two areas of
tal factors arepoorly understood. known Petrogalepenicillata habitat on the
Extensive wildfires in 2003 provided an northern slopes of Little River (37° 06’ S,
opportunity to study the effects of fire on 148° 20’ E) and Fann Creek (a tributary of
dry Acacia scrub communities on Lower Little River, 37° 07’ S, 148° 20’ E).
Devonian volcanics of the Little River Vegetation was dominated by tall Acacia
Gorge in East Gippsland, Victoria. We scrubon dryslopesand spurs, withE. alhens
undertook stratified floristic surveys to woodlands on locallycooler(eastern) aspects,
assess short-term impacts of mild and and shrubland dominated by P. trachyphylla
intense fire on understorey composition on exposed rocky outcrops. Some intergrad-
and vigour, and to establish benchmarks ingoccurred amongthese communities, espe-
for monitoring vegetation response to fires ciallyatFarmCreek.
over a longer time frame. Surveys were Sampling was undertaken in November
also relevant to management of habitat of 2004 to January 2005, and was confined to
the Brush-tailed Rock Wallaby Petrogale vegetation with a prominent Acacia over-
penicillata (.IE Gray 1825), as the Little storey. Quadrats were 10 x 10 m, and per-
River Gorge area provides the last remain- manentlymarkedto allow long-term moni-
ingrefuge forthis species in Victoria. toring. Although the study was undertaken
almost two years after the Bogong
The StudyArea Complex fires, a range of fire intensities
The Little River, upstream of its junction was easily distinguishable:
with the Snowy River in East Gippsland, • Sites with deep, non-charred, ground-
forms Victoria’s deepest gorge. The area is layer litter and occasional live shrubs
highly dissected, with steep slopes and deep with sparse to abundant twig-lichens
valleys ranging in elevation from 200 to [Usnea sp.), wereclassed as unburnt.
1000 m A.S.L., and occupies a rain shadow • Sites with visible charcoal and partially
mm
with an average ofaround 600 rainfall burnt twigs, llre-scars on trees and larger
per annum. Soils are derived from Lower shrubs, no twig-lichens, and <20% kill of
Devonian volcanics (rhyodacites), and con- theAcacia canopy were classed as lightly
trast with the coarse grained granodiorites burnt.
and UpperOrdovician sediments(siltstones. • Sites with >80% kill of the overstorey
schists, slates, shales and claystones) occur- Acacia trees were classed as intenselv
ring immediately to the north and east burnt.
(Clayton-Green andWimbush 1988). At each site we aimed to sample four
Typical vegetation ofthe local region is replicates each of unburnt, lightly burnt
grassy woodland dominated by White Box and intensely burnt Acacia scrub.
Eucalyptus alhens, grading to other euca- However, no unburnt sites were found at
lypt forests and woodlands at higher alti- Little River, so the final sampling design
tudes and on poorer soils. Dry Acacia comprised four unbumt plots (Farm Creek
scrubs occuras scatteredpatches associated only), eight lightly burnt plots and eight
with steep rocky valleys and slopes intenselyburntplots (both sites).
amongst these vegetation types, both on Plots were placed so as to sample the
rhyodacites and sediments, but rarely on rangeoftopographicclassespresentateach
granodiorites (Clayton-Greene and site (NE to NW slopes, drainage lines and
Wimbush 1988). rocky spurs). Within each topographic
Large areas of the Little River valley class, we sought to place plots ofeach fire
were burnt in the extensive Bogong intensityclasswithin close proximity, lead-
Complex wildfires during .lanuary 2003. ing to paired (lightly-burnt and intensely-
Fire intensity varied depending on aspect. burnt) plotsat Little River, and sets ofthree
Vol. 124 (6) 2007 321
Research Report
(unburnt, lightly-burnt and intensely-burnt sional solutions, compared with minimal
plots) at Farm Creek. Plots were randomly improvement in higher dimensions, led to
placedwithintheseconstraints. selection ofthe three-dimensional solution
for further exploration. Stress values of
Floristicmonitoring
Percentage foliage projective cover was the selected solutions were 0.1 1 for the
presence-absence and 0.10 for the quanti-
visually estimated for all higher plant and
tative analysis.
<rfeo1rcn%ksvcpioesvcueiarel.scaoSnvpdeercfioweresrbeatrhaealtlgorwcoeaurtneedd,parleintsoteemrnitnaanaldt wiCtohrrtehleatdiiosntsriobfuteionnviorfonsmietenstaolnatthteribfuitnaels
cover of0.5%. This may have led to over- eoradcihnadtiiroencstlwyeornetoextahmeinoreddinbaytipolnotatnidn,g
eofstteinmaptreesseonftattotvalercyolveorw,coavsers.peAcireasngweeroef wofhemraexaipmpruompricaotrer,elabtyiocnal(cRumlaaxt)inugsivnegctotrhes
other variables (richness and cumulative vector-fitting procedure of DECODA.
cover for native and shrub cover including
and excluding P. trachyphylla. forb cover, Monte-Carlo tests, using 1000 random per-
fgrraosmspcroivmearraynddateax.oticcover)werecalculated mmuattaettihoenssiogfnitfhiecandcatea,ofweearceh uvescetdortocoersrtei--
lation. For the purposes ofvector correla-
Dataanalysis tions, topographic class was treated as a
Richness and cover estimates were sequence fromgullyto slopetorocky spur,
analysed in GenStat (2003), using general and aspect a sequence from north-west to
linear regression to perform analysis of north-east. Where appropriate, vectors
variance ofthe unbalanced 8 block (pairor were used to order sites and species into
triplet) X 3 fire intensity design. It was two way tables indicating species con-
assumed that effects offire intensity were tributingtorelevanttrends.
consistent across sites (i.e. that there was Results
no sitex fire intensity interaction), and that Acaciascrubs ofthe studyareawere domi-
fire intensity and pre-fire vegetation com-
nated by CtirrawangA. doratoxylon, which
position were not significantly confounded. formed small trees to 20 m on lower
For each variable showing a significant slopes, and became smaller (to 3 m) on
overall effect, treatment means adjusted
upper slopes and rocky spurs. A. silvestris
pfoarrbeldocukseifnfgectpsowsetr-ehoccalctuelsattsed(aFnidshceorm’-s wdaosmioncacnastioonnalthateLritotclkeieRsivters,itaensdatbeFcaarmme
Protected Least Significant Differences,
Creek. Other overstorey trees included
Steel andTorrie 1981).
patches ofE. alhens, emergent Eucalyptus
Ordination was used to explore patterns
in understorey composition. Two ordina- sEauxcaatliylpitsusatsmLiitthtilieaRtiFvaerrmaCnrdeemka,lalnede-ofcocram-
tions were performed and compared, one
sional Allocasuaritia verticillata at Farm
using presence-absence floristic data and Creek. On very rocky sites P. trachyphylla
one using quantitative floristic data. For
replaced A. doratoxylon as the dominant,
each analysis, floristic data (excluding tree though usuallywith scatteredemergentaca-
species) were used to produce a distance cias. The understorey ofthe Acacia scrubs
matrix usingthe Bray-Curtis coefficientof was generally sparse beneath the intact
driicssimmuillatriidtiym(eFnaistihonetalal.sc1a9l87i)n.gN(onn-MmDeSt)- caAnotpoyta(le.go.fF6ig4.nlaa)t.ive plant species was
analyses were performed on the distance
recorded across the 20 plots (39 at Little
matrices using the software package
DECODA River and 54 at Farm Creek), including
(Minchin 1989), and results
were transferred to PC-ORD (McCune asenvdensugbrasssh,ruobn,etsweodgev,in3e0afnodrb,se1v7enshtrruebe
and Mefford 1999) for plotting and visual species. A high proportion (20%) ofthese
analysis. Analyses were performed in one
species are classified as rare or vulnerable
atondfivgelodbiamlennsMioDnSs.usCionngsi10derraanbdloemssttraerstss ainndViEcntovriiraon(mDeenptar2t0m0e5n)t.oOfnSluystsaiixnaebxioltiitcy
reduction from the two to the three dimen- species were recorded, all ofwhich were
322 The Victorian Naturalist
Research Report
herbaceous. Native species richness per 10
X 10 m plot ranged from 3 to 22 species,
and exotic richness from 0 to 4. Total
(cumulative) native cover was highly vari-
able (range 6-71%), while exotic species
cover was consistently very low (<2%).
Understorey composition ranged from
grass to shrub dominated and was generally
different from surrounding grassy wood-
landsanddryforests.
Unburnt plots had a dense litter layer of
Acacia phyllodes and twigs, with small
amounts ofbare soil and variable amounts
of outcropping rock. Lightly burnt plots
had approximately halfthe litter cover of
unbumt plots (comprising litteraccumulated
since the fire and/or incomplete combus-
tion ofthe pre-fire litter layer), and signifi-
cantly more bare soil. On intensely burnt
plots, the litter layer was largely removed
by the fire and had not been replaced
owing to canopy death. Bare soil levels of
intenselyburnt plots were similarto lightly
burntplots (Table I).
Understorey cover at tw'o years post-fire
was significantly and dramatically infin-
enced by fire intensity. Unburnt and lightly
burnt plots generally had a sparse under-
storey of scattered shrubs and grass tus-
socks, with the exception ofone unburnt
plot located on arockyspurandwith avery
dense cover (70%) ofthe shrub Philotbeca
trachyphylla (Fig la and b). Understorey
cover increased to an average of35% on
intensely burnt plots (Fig. Ic). an increase
ofover 400% (excluding the anomalous P.
trachyphylla plot. Table 1 ). Increases in
cover due to intense burns were significant
for the native shrub, grass and forb compo-
nents ofthe understorey, but there were no
significant effects on weed cover. Native
shrub richness also increased on intensely
burntplots (Table I).
Despite considerable variation in under-
storey composition between differing sets
or pairs of plots, fire intensity correlated
significantly with the nMDS ordinations
(Figs. 2a. b; Table 2). On the presence-
absence ordination, the correlation was rel-
atively weak (R|„ax“ 0-65). Unburnt plots
were usually separated from burnt plots,
wbuetrelipgohotlrylybudrinsttianngudisihnetdenfserloymbeuarcnht optlhoetsr AFicga.ci1.aTsycpriucbals raatngLeittilneunRdievresrtoirneyredleantsiiotnytion
(Fig. 2a). On the quantitative ordination. burning intensity a) unburnt, b) lightly burnt,
c) intenselyburnt
Vol. 123 (6) 2006 323
Research Report
Table 1. Adjusted means for understorey and ground cover characteristics Acacia doraloxylon
scrub twoyears afterburningat different fire intensities. Means with different superscripts are sig-
nificantlydifferentfromeachother(P<0.05).
Unburnt Lightly Intensely P
Burnt Burnt
Baresoil 6-0^ 27-9'’ 244b 0.01
Exposedrock 360^ 37-9“ 26-6“ ns
Litter 49.3a 274!“ 1-(R <0.001
Nativerichness 114“ 12-5“ 15-34“ ns
Nativecover(%) 254“ 8-3'“ 35-5“ 0.021
Nativecover(%)excl.Philolheca 81“ 7.7a 34-7“ <0.001
Shrubrichness 2-2“ 3.3a 4.9b 0.006
Shrubcover(%) 21-8“ 14“ 18-6“ ns
Shrubcover(%)excl.Philotheca 4-5“ 0-9“ 17.9b 0.003
Grassrichness 14“ 1-6“ ns
GFroarsbsrcicohvneerss(%) -07--01““ 27-.09“a 190--91“*’ 0n.s002
Forbcover(%) 3-6“ 4.9a 7.3b 0.006
Weedrichness 04“ 0-6“ 1-4“ ns
Weedcover(%) 0-2“ 0-3“ 0-7“ ns
*negativevalueduetoadjustment;actualmean I-9
Table 2. Vectorcorrelations (R„a,) ofplotcharacteristics with ordinations using presence-absence
andabundancedata. Pindicatessignificanceofcorrelations.
Presence/Absence Abundance
Plotcharacteristics Rmax P '^max P
Fireintensity 0-65 0-031 0-87 0-000
Topography 0-74 0-002 0-69 0-019
Site 0-62 0-064 0-66 0-021
Aspect 0-23 0-823 0-15 0-930
(a) Q unburnt
# lightlyburnt
intenselyburnt
Axis1
Fig. 2. nMDS ordination results for (a) presence-absence data showing relationships with burning
intensity.Axesshowingmostdistinctdifferencesarepresented.
324 The Victorian Naturalist
Research Report
(b) O Burn
• onburnt
# lnigthetnlsyebluyrnbturnt
Philolhecathicket_
CN
(A
<
o
o
o
Axis1
Burn
O
unburnt
# lightlyburnt
intenselyburnt
O
oPhilothecathicket
co
Xc»
<
Axis2
Fig.2.cntd.nMDSordinationresultsfor(b)abundancedata(axes I versusaxes2andaxes2versus
axes3,showingrelationshipswithburningintensity..
Vol. 123 (6) 2006 325
Research Report
Ri„„^ increased to 0.87 and intensely burnt rocky spurs (Table 4). There was little
plots were distinctly separated from all effect ofaspect within these topographic
others (Fig. 2b). This suggests that abun- classes, probably reflecting the limited
dance of species played an important role range of aspects occupied by Acacia
in distinguishing lightly burnt from scrubs (NE toNW,Table2).
intensely burnt plots, consistent with There was also some distinction in under-
analysesofcoverestimates. storey composition between the two sites
Differences in floristic composition related (significant on quantitative ordination only.
to fire intensity arc indicated in Table 3. A Table 2). This separation could be attrib-
numberofspecies, suchasPlaty’sacelanceo- uted to the absence from the Little River
lata, E. albens, Oxalis peremians, Galium plots ofspecies such as Gonocarpus elalus
liratiim and Dichondra repens showed no (present at Little River, but not in the
detectable relationship. Few understorey plots), Cheilanthessieberiand Ozothumnus
species were clearly more prominent in thyrsoideus, and absence from the Farm
unbumtorlightly burnt plots, but it is possi- Creek plots of Cynoglossum australe,
ble thatP. IrachyphyllaandDkmellare\’olu- Parietaria debilis and Cyphanthera albi-
ta were disadvantaged by intense fires. As cans, as well as a greater frequency of
well, the overstorey species AUocastiarina species such as Dodonea rhombifolia,
verticillata andA. doratoxylon were less fre- Senecio hispidulus and Nicotiana suave-
quent or abundant in intensely burnt plots olensat Little River,
owing to canopy death, consistent with our
Discussion
samplingstrategy,
A distinct suite of species were absent Acacia doratoxylon has not previously
from the understorey of unburnt plots, been described as dominant in Acacia
sparseorsporadic in lightly burntplots, and scrubs in Victoria, but A. doratoxylon
scrubs are noted from the adjoining
D(frToeadqbuolenenate3o)ar.arbThuhonemdbsaienftoilinnicail,ntuedMneisdieleythhlbeeurnnsbthercplukoibtuss CBlyaaydtboonw-iGlrdeerenneessaonfdNWSiWmb(uKseihth(21090848)).
dicliua, Cyphanthera albicans (absent from sampled a range ofAcacia scrubs in the
Farm Creek), Solantim linearifolium, and Upper Snowy region, and found A. sil-
the forbs Nicotiana suaveolens and vestris was dominant in nine of 11 stands.
Sigesbeckia aiistraliensis. A number of However, they did not sample scrubs on
rhyodacites owing to poor accessibility.
otherspecies occurredoccasionallyorinfre-
Further surveys to elucidate the extent,
quently in unburnt plots, but increased
markedly in frequency or abundance with status and environmental determinants of
increasingfire intensity, e.g. Notodanthonia A. doratoxylon scrubs in Victoria would
thus be ofvalue.
longifolia, Einadia hastala. Solariumprino-
phyllum and Lepidium pseudotasmanicum. The structure and fioristic composition of
Cynoglossum australe and Senecio hispidu- A. doratoxylon compared with otherAcacia
lus increased inburntplots but with noclear scrubs is also ofinterest. In our study area.
A. doratoxylon formed low forests to tall
effectsoffireintensity.
shrublands with open understorey, but did
Ordinations using either presence-
not form dense, impenetrable thickets as
absence or abundance data also correlated
commonly reported forA. silvestris scrubs
significantly with topography (Table 2), (Clayton-Greene and Wimbush 1988). A
with all five samples from rocky spurs dis- high proportion ofthe component plant
tinguished from other topographic classes
(slopes and drainage lines). A range of species recorded in this study are listed as
species (e.g. Goodenia macmillanii, rare or vulnerable in Victoria (Department
Lepidosperma laterals, NicoHana suave- ofSustainability and Environment 2005),
olens, Arthropodium rnilleflorum, tbyupticitalisonfototkhneorwtnypwehsetohfeArctahceisae sspcercuibe.sWaree
Dichondra repens, Cheilanthessieberi and
Sigesbeckia aiistraliensis) was absent from also noted relatively sparse and slow-grow-
ing regeneration of doratoxylon (e.g. live
plots on rocky spurs, while Isotoma axil-
cover values in intensely burnt plots were
plahryils,laA.wesirlevesmtroisstanfdrePqhuielnotthiencapltortaschoyn- usually<1%,withseedlingsmostlylessthan
326 The Victorian Naturalist
Research Report
Table 3. Two way table indicating the influence offire intensity (U=unburnt, L=lightly burnt,
^intensely burnt)on understoreycomposition in 20AcaciascrubsamplesatLittle RiverGorgeand
Farm Creek. Sites were ordered using the vector ofmaximum correlation for fire intensity for the
quantitative nMDS ordination, and species were ordered according to their weighted average loca-
tionsalongthisvector. Percentabundance is represented bythe followingclasses; I,<1%; 2, 1-2%;
3,3-5%;4, 6-10%; 5, 11-15%; 6, 16-25%; 7,26-50%; 8, 51-75%;9, 76-100%. Allspeciesthatwere
recorded from 3 or more plots are included. Note that tree species (bold text) were not included in
ordinationanalyses; lowtreeabundanceon intenselyburntplotsisduetocanopydeath.
LUULULULLLLL lllllll
Fireintensity 1
-8I2---22--- 3---1--
Philolhecatrachyphylla --3-I1I---2- -------
Allocasiiarinaverticillala -243----I--- 1-1--12
AcaciasUvestris 868898777887 1111211 I
Acaciacloratoxylon --1-11---11- ------- 1
Dianeiiarevolutas.l. 1---111----- -]-----
Lepidospermatalerale I----11-----
Clemalismicropinila I----11----- ---]---
GaliumhinifoHum --I-1-----1- |- -- --
AGuosntorcoasrUppuasdeelnasti(ifslora ----2-----1- --- 1l--.-.-----
OzolhammisIhyrsoideus l-ll-l11l-l--- --111--
PClaaltaynsdarcienilaanccaeloylpaitraata 414-3--131--6---3-2-21-- 6- 2I-32-5-1--1
Eucalyptusaibens -----11----- ----]-- 2
Ptewosorusmtifolius I--I---1-1-- ---1--1
Crasstilasieheriana l-l-lll-llll llllll-
Oxalispereimans --i--||----. --ll|--
Arthropodiummilleflonim I-I----II1-- 1-1--1-
Galiumliratum --II--- ---. ------l 1
Isotomaaxillaris 1-1--II-1-1-- --1211- l
Dichondrarepens ----1II---I- -21----
Cheilanthessieheri -----11I1--I l-llll-
Lepidiumpseiidolasmanicum --1----1---- ii-|l--
Solantimprinophylltim 1111-22-22-3 4-22321
Einadiahastata --1--1--11-- --11--I 1
Sonchusoleraceus --1--1 ------ I
Sonchusasper -- ----l-lll-l ---1-11- 1
Seneciohispidtihis -------1---- 221211- 2
Solam/mlinearifoliiim 111121112231 4642552
NolodanlhonialongifoUa ---------1-1 ---2-1- 3
Cynoglossmnausirale ---- 2-2 --3321- 1
Nicolianasuaveolens 1-- ---1-11 I
Parietariadebilis I----I1----2 ---I5--
SGiogoedsehneickaimaaacumsiilrlaalniienisis .- --I--------1!---- -.---1I2 -I-32- 1
Cyphaniheraalbicans ---1---I11I- 2134343 2
Muehlenbeckiadiclina ------- I--I --24-26 5
DodonaearhombifoUa 1 ------3 6
EucalyptussuxatUis 3
0-5 m), compared with dense and robust al. 1982, Clayton-Greene and Wimbush
regeneration in nearby A. sUvestris scrubs 1988, Keith and Bedward 1999). Richness
(post-fire live cover to 50%, with many ranged from 3-19 species (average 12) in
seedlingsto2m inheight,pers. obs.). unburnt plots; higher than reported by
Keith (2004) torAcaciascrubs in NSW.
Effects offtreon vegetationstruciureand Surveys indicated that light burns led to
Jlofistics minimal change in total understorey abun-
Consistent with earlier reports, the under- dance, and sporadic establishment of a
storey of the Acacia scrubs ofthis study range ofspecies that were rare orabsent in
was sparse beneath the intact canopy (e.g. the understoreys ofunburnt sites. By con-
Fig. la), ordominated bydensePhilotheca trast, burnsthatwere intenseenough to kill
trachyphylla in rocky situations (Forbes et the Acacia overstorey stimulated dense
Vol. 123 (6) 2006 327
ResearchReport
Table4.Twow'aytable indicatingtheinfluenceoftopography(G=gully, S=slope,R=rockyspur)on
underslorey composition in 20Acacia scrub samples at Little River Gorge and Farm Creek. Sites
were ordered using the vector ofmaximum correlation for topography for the quantitative nMDS
ordination,andspecieswereorderedaccordingtotheirweightedaverage locationsalongthisvector.
Percent abundance is represented by the followingclasses: 1, <1%; 2, 1-2%; 3, 3-5%; 4, 6-10%; 5,
11-15%; 6, 16-25%; 7, 26-50%; 8, 51-75%; 9, 76-100%. All species that were recorded from 3 or
moreplotsareincluded.
Topography G S S S S GS S S S S S S S R S R R R R
Pleiirosorusrutifolius - 1 1 1
Goodeniamacmillanii 2 1 1 5 - 1 1
Ozotbanmusthvrsoideus - 1 1 1
Clematismicrophvlla - 1 1 - - 1 1 - - - - - - - - - - - - -
Galiumhimfolimn - 1 1 - - 1 1
Lepidospermalaterale - 1 1 - 1 1 1
Nicotianasuaveolens 2 - - 2 3 3 - 1 2 - - - 1 - - - - - - -
Arthropodiurnmilleflonmi - 1 1 1 1 1 - - 1
Dichondrarepens - 1 1 1 1 2 1 1 1 - - - - - - 1 - - - -
Cheilanthessieberi - 1 1 - 1 - - - - 2 1 1
Cvnoglosswnaustrale 1 - - - - 2 - 1 1 - - - 1 - - - - - - -
Sigesbeckiaauslraiiensis - - - 1 1 2 - 1 1
Oxalisperennans 1 1 1 1 1 1 1 1 1 1 1 1 - 1 - 1 1 - - -
Lepidiwnpseudofaswanicnm 1 1 1 1 1 1 - 1 1 1 1
Sonchnsasper 1 1 1
Calandriniacalvptrata - 1 1 1 1 1 1 - - - - - - 1 - 1 - - 1 -
Notodantbonialongifolia 1 1 1 5 4 2 1 5 2 6 2 3 3 2 2 1 4 1 1 1
Einadiabastata 3 2 2 3 2 2 1 2 2 - - - 1 2 1 1 4 - 1 1
DianellarevoJuta - - 1 ~ “ 1 1 1 - 1
Aiistrostipadensiflora 1 2 1 : :
Parietariadebilis - - - - - 1 - 1 1 - - - - - 1 - - - -
Seneciohispidiihis 1 - 1 - - 1 - 1 1 - - - 2 1 1 - - 1 - -
Sonchnsoleraceus - - 1 - 1 1 - - 1 - - - 1 1 I 1 - - - -
Muehlenbeckiadiclina - - - 3 3 4 - 4 1 1 - 1 5 1 3 - 2 1 1 -
Gonocarpuselatus 1 1 1 - - - 1 - - - -
Acaciadoratoxylon 7 7 8 2 1 1 8 1 8 1 9 8 1 7 1 8 1 7 8 6
Solamim linearifolium - - 1 1 2 - 1 - 2 - - - - - - 2 1 - -
Cypbantberaalbicans - - - - - 1 - 3 1 - - - 2 - 2 - - - 1 -
AUocasuarinaverticiHata - 1 1 - - - - - - - 1 2 - - - 3 - - - -
Eucalyptusalbens - 6 - 5 3 2 4 1 2 2 3 2 2 3 - 3 6 - - 4
Dodonaearbombifolia 1 - - - 2 4 - 2 - - - - 1 6 - - 1 - "
Platvsacelanceolata - 1 1 1 1 - - 1 - - - 1 -
Solannmprinopbvlhim 1 1 - 1 - 1 1 1 - -
Galiumliratum - - - - 1 - 1 1 1 - - - 1 1 - 1 1 1 - -
Eucalyptussaxatilis - - - - - - 1 - - - - - 3 - 3 - - - - -
Crassulasieberiana - - - - - 1 1 - 1 - - - - - 1 - - 1 1 -
Isotomaaxillaris 1 - 1 1 - 1 1 -
Acaciasilvestris 1 1 1 1 2 4 1 - 3 2
Philothecatrachyphylla 1 2 - 1 3 2 2 8
establishmentandrobustgrowth ofspecies with the breakdown of root systems of
absent from understoreys ofunbumt plots, deadAcacias (FisherandBinkley2000).
leading to a significant increase in native Theimportanceoftheoverstoreyforsup-
shrub richness and understorey abundance. pressing understorey species is further evi-
As well, intense burns promoted lush dent from field observations, where sharp
growth ofthegrassNotadanthonia longifo- boundaries between dense and sparse
liaand forbs suchasEinadiahastata. understorey are coincidentwith boundaries
Processes leading to these understorey between killed and live canopy respective-
changes are likely to involve both direct ly(e.g. Fig. 3).
effects offire on seed dormancy and ger- Infrequent, intense fires have previously
mination, and release from intense compe- been recognized as important for stimulat-
tition for light, soil moisture and/or soil ing recruitment ofoverstorey species and
nutrients after overstorey death. As well, a thus the perpetuation ofAcacia scrubs in
flush ofsoil nutrients may have occurred farsouth-easternAustralia(Clayton-Greene
328 The Victorian Naturalist
Research Report
and Wimbush 1988). This study indicated urally somewhat denser than in unburnt
that a range ofunderstorey species are also areas. Nevertheless, observations ofthe
dependent on intense fires in these commu- vegetation before the fire (R Martin, pers.
nities, where they oecur mostly as post-fire comm.) confirm that many species that
ephemerals (sensu Gill 1993, becoming flourished after the fire (e.g. Cyphanthera
abundant only after intense fires, and exist- albicans, Muehlenheckia diclina,
ing as scattered individuals and/or a dor- Dodonaea rhombifolia) were uncommon
mant seed bank for long periods). Many of to very rare in the understorey before the
these species in Little River Gorge are vul- fire, and that other species were common
nerable, rare or disjunct in Victoria (Table but less robust and abundant than after the
5), and thus the appropriate fire manage- fire(e.g. Goodeniamacmillanii).
mentofAcacia scrubs is ofsignificantcon- Implicationsforfiremanagement
servation interest. Some preliminary conclusions regarding
While the dramatic effects of canopy
burns on understorey composition and appropriate fire regimes for managing
Acacia scrubs and Petrogale penicillata
vigour were readily evident in the field, a
caveat must be placed on conclusions of habitatin Little RiverGorgemightbedrawn
this study. Because monitoring plots were basedon conditionsattwoyearspost-fire. In
established after the fire, some confound- particular, the strong differentiation of
ing of understorey composition with fire intensely-burntplots from unbtimtand light-
intensity may have occurred due to the ly-burnt plots suggests that only intense,
influence of pre-fire vegetation on fire canopy-killingfiresaresuitableforreleasing
behaviour. For example, observations many fire-recruiting species from the seed-
bseugfgoersettthheatfisroem(eRarMeaarstihna,dpheirgsh. ccoovmerm.o)f bstaonrke,yagnrdowftohripnrAo.modtoirnatgoxvyilgoonrosucsrubusn.deBry-
dmroyrgerasfsltahmatmcah,blaendtthheasne moatyhehravearbeeaes.n cuosnutarlalsyt,apcpolniterdolilnedm,alnoawg-einmteenntsibtyumfisremsaays
vCeognesteaqtuieonntilny,intietnisselpyosbsuirbnlteartehaastwparse-fniart-e nportomboetseufsfiigcniiefnitctaontkirllectrhueictmaennotpyanadndgrhoewntche
ofunderstoreyspecies.
Fig. 3. Sharp boundaries in underslorcy abundance relating to canopy death were typical across
burntareasat Little Riverand Farm Creek. All ofthe area in this image was burntduring the 2003
lire. However, canopy on the left was killed, corresponding with lush underslorcy growth, while
canopyontherightwaslargelyintactandunderstoreyremainssparse.
Vol. 123 (6) 2006 329