ltant engaged within the program. This article pro-
vides an overview of 4 years’ treatment of Bitou
Bush and the effects of this treatment on the
recovery of native vegetation.
Introduction
O
ver the Christmas and New Yearperiod of 2001/2002, an intense
wildfire burnt through approximatelytwo-thirds of Bundjalung National Park,including its extensive dune system(Fig. 1). Mapping of the nationally signi-ficant weed Bitou Bush (
Chrysanthe-moides monilifera
ssp.
rotundata
)
(Box 1 and Fig. 2) had shown that in2002, about 75% of the foredunes in thepark contained heavy infestations (i.e.greater than 40% of total vegetationcover) (NPWS 2001; Fig. 3). Based onthe high level of Bitou Bush infestationof the dunes and understanding of therole of fire as an important componentin Bitou Bush control (Weiss 1983; Flower& Clarke 2002), a proposal was developed
Figure 1. Prior to 2002 about 75% of the foredunes in Bundjalung National Park containedheavy infestations of Bitou Bush (the lighter-coloured vegetation visible along the coast). This is justa portion of the approximately 900 km of the New South Wales coastline (80% of the coast) alreadyinfested by Bitou Bush. Without successful control, this nationally declared environmental weed isconsidered to have potential to expand its range across a much larger area of the Australiancoastline. (Photo: Department of Environment and Conservation)
emr_262.fm Page 79 Wednesday, June 28, 2006 1:22 PM
Box 1. Bitou Bush (Chrysanthemoides monilifera ssp. rotundata)
Bitou Bush is listed as one of the
initial 20 Weeds of National Signi-
ficance (WONS – see ARMCANZ,
ANZELL and ANZFM 2000) and is
a declared noxious weed. Invasion of
native plant communities by Bitou
Bush is listed as a Key Threatening
Process under the TSC Act 1995.
Mapping of Bitou Bush in 2000
found approximately 900 km (80%)
of the New South Wales coastline
infested, with a conservative esti-
mate of 36 000 ha infested (NPWS
2001). Bitou Bush originates from
South Africa and was introduced
into Australia in 1908.
In the 1950s, Bitou Bush was
among a number of plants trialled
for sand-dune stabilization and was
still being recommended for planting
in 1970. A mature Bitou Bush plant can produce up to 48 000 seeds per plant per year (Weiss et al. 1998). Prostrate stems
enable vegetative reproduction to occur by sending roots from adventitious buds. Seeds can be spread over large distances by
animals such as birds and foxes. Dispersal of seeds also occurs by wind, water, vehicle movements, and movement of contaminated
soil. In established infestations, soil seed banks can contain up to 5000 seeds/m2, with a small percentage of seeds remaining viable
in the soil for up to 10 years, thereby providing a source for reinfestation. Disturbance, especially by fire, promotes seed germination,
but germination still occurs in undisturbed situations. A variety of controls including biological, chemical and physical methods are
available to control Bitou Bush with site-specific assessment required to determine the most appropriate technique or combination
of techniques for a particular location.
In recent years, national, state, regional and local strategies have been prepared and are being implemented to reduce the
impacts of this plant. Many coastal community ‘care groups’ focus their attention on this plant. Successful control programs
share a number of common features:
• A variety of stakeholders working cooperatively, often with a formal plan that outlines priorities, roles and responsibilities
• A commitment from decision-makers at all levels
• Persistence and continuity for long-term commitment
• The presence of key individuals who coordinate and motivate participants
• The matching of the scale of work (including follow up) to the available resources
Figure 2. Bitou Bush is listed as one of the initial 20 Weeds of National Significance (WONS)and is a declared noxious weed. Invasion of native plant communities by Bitou Bush is listed as aKey Threatening Process under the NSW TSC Act 1995. Bitou Bush, originated in South Africa andwas introduced into Australia in 1908. (Photo: T. McDonald)
emr_262.fm Page 80 Wednesday, June 28, 2006 1:22 PM
by the Department of Environmentand Conservation (New South Wales)to take advantage of the fire and controlBitou Bush to prevent it from completelydominating and degrading the entiredune system (80% or more cover).
Prior to the fire (the extent of whichis mapped in Fig. 3), successful butlimited Bitou Bush control had beenundertaken at a small number of siteswith community support. The DefenceDepartment, which uses a large area ofthe park as a bombing range, agreed toparticipate in this program and providesubstantial resources.
S i te descr ipt ion
Declared in 1980, Bundjalung NationalPark, is a large (20 326 ha) conservationreserve on the subtropical northernNew South Wales coast (Fig. 3). Itextends over approximately 35 km ofcoastline from south of Evans HeadVillage to the Clarence River. Togetherwith the adjoining World Heritage-listedIluka Nature Reserve (136 ha) to thesouth and the 3800-ha BroadwaterNational Park to the north, the threereserves form the second largestprotected area on the New South Walesnorthern coastline. A 10 655-ha area ofthe park is gazetted as wildernessunder the Wilderness Act 1987.
Bundjalung National Park contains avariety of coastal vegetation communi-ties including wet and dry heathland,littoral rainforest, mangrove, eucalyptforests and
Melaleuca
woodlands, aswell as freshwater and estuarine wet-lands, headland grasslands and dunecommunities (Griffith 1983). A total of712 plant species have been recordedin the park, including 47 exotic species(NPWS Wildlife Atlas database, 2006).In Bundjalung NP, as in most of theNorth Coast, Bitou Bush occurs mainlyin the frontal dune landform but alsoextends to hind dunes and duneswales. This dunal landform is naturallysusceptible to Bitou Bush invasion. Thepredominant vegetation type on thefrontal dune today is Coast Banksia(
Banksia integrifolia
) woodland; andremnants of littoral rainforest, a Threat-ened Ecological Community in New
South Wales, occur in several places inthe 35-km-long dune system. (Plantnomenclature used in this note is con-sistent with Harden (1990–1993)).
Aboriginal people have lived in thearea for thousands of years and havecontinued to maintain strong links withthe park. Impacts from this period onthe dune are likely to have been minimal.However, European uses of the areasince settlement have caused significantimpacts primarily through mineralexploitation and defence activities.For example, gold was mined in thelate 19th century and mineral sand-mining occurred over extensive areasof dunes and heathlands between the1930s to the early 1980s. Quarrying forgravel and rock occurred in severallocations and there was sporadic andseasonal cattle grazing throughoutmuch of the area. Commercial and re-creational fishing have a long traditionand the park has always been popularwith day visitors and campers. Twotarget areas occur in the north of thepark and are included within the2572-ha Evans Head Air WeaponsRange (known as the Bombing Range).These have been used by the defenceforces since the 1940s with the south-ern part of the range still operationaland used for various military purposes.
Although all these disturbances havehad some effect on the dunes, the high-est impact on the dunal system in thesouthern section has arisen from sand-mining. This extensively disturbed thedune landforms, reshaping the dunesto a much lower profile and leaving apatchwork of cleared and uncleared/mined remnant vegetation (includingpatches of littoral rainforest, which wasprobably much more extensive in thesouthern section than is the case today).Post-mining rehabilitation in the south-ern section involved revegetation witha limited suite of species includingCoast Banksia, Coastal Wattle (
Acaciasophorae
) and Sand Spinifex (
Spinifexhirsuta
). In the northern section, in con-trast, dunes were not significantly dis-turbed by sandmining but certain areaswere subject to disturbance for recrea-tional access by vehicular and foot traffic.
For geographical and administrativereasons the park is managed by boththe Northern Rivers and North CoastRegions of the Parks and Wildlife Divi-sion of the Department of Environmentand Conservation, (NSW) formerlyknown as the NSW National Parks andWildlife Service. This arrangement workswell in practice, as there are distinctpark management issues in the differ-ent regions, as well as issues requiringa consistent approach throughout thepark (e.g. fire and weed management).
Bitou Bush invasion in Bundjalung National Park
At the commencement of the project,Bitou Bush was the most abundantweed in Bundjalung National Park. Itreached greatest densities in previousforedune sandmined areas and can alsobe found in areas up to 5 km inland,remote from camping or residentiallocations, where it can be the onlyweed species present.
With growing concern about theeffect of this weed upon native vegeta-tion (Box 2 and Fig. 4), the density andextent of infestation of Bitou Bush wasmapped along the New South Walescoast in 2000 as part of the preparationof the New South Wales and NSWNorth Coast Bitou Bush managementstrategies (NPWS 2001; Scanlon 2001).The survey, based upon aerial photo-graphy interpretation and ground truth-ing, found that Bitou Bush infestedapproximately 680 ha of BundjalungNational Park, occurring in a nearlycontinuous strip from north to south(Fig. 3). Within this infestation 320 hawere mapped as heavy (>40% cover),107 ha as medium (10–40%) and 235 haas light (<10%).
Control of Bitou Bush in the park
The first attempt at larger scale controlwas the release of biological controlagents, the Bitou Bush Tip Moth(
Comostolopsis germana
) in 1990 andthe Bitou Bush Seed Fly (
Mesoclanispolana
) in 1996. Until this project
emr_262.fm Page 81 Wednesday, June 28, 2006 1:22 PM
Figure 3. Location map of Bundjalung National Park, showing Bitou Bush distribution prior to treatment. As in most of the North Coast, Bitou Bushin the park occurs mainly in the frontal dune landform, extending inland where conditions are suitable. Of the approximately 680 ha affected, nearly halfwas mapped as heavily infested. (Map: Department of Environment and Conservation)
emr_262.fm Page 82 Wednesday, June 28, 2006 1:22 PM
commenced, Bitou Bush control workwas generally confined to visitorarea precincts or as part of morecomprehensive restoration projects onsmall sites, often involving communitygroups. For example, in 1995, a jointcommunity group (Iluka LandcareGroup) and NPWS project began torestore the natural vegetation of the5-ha Iluka Bluff area at the southernend of the park. A variety of controls(including overspraying, fire, mechanicalcrushing and manual follow-up methods)were used in this iconic and heavilyinfested headland and dune area with
extensive replantings in certain sections.(CRC for Australian Weed Management2003; Natural Heritage Trust 2003).
Cost-effective control of large BitouBush infestations has been available sincethe early 1990s through the developmentof aerial herbicide application techniquesby the former NSW Agriculture (Toth
et al
. 1991, 1993, 1996). Low volumeapplication of principally glyphosate-based herbicides are applied in winterby helicopter at rates sufficient to killBitou Bush and cause negligible impactson most native plants that are found inBitou Bush-infested areas.
Project in i t iat ion and planning
The aim of the long-term Bitou Bushcontrol program currently underwayin Bundjalung National Park is toprotect biodiversity by the targetedpost-wildfire control of an extensiveBitou Bush infestation in the coastaldunes, to create conditions suitable forthe regeneration and restoration of thenative coastal vegetation (McDonald2002).
An opportunity to gain efficienciesfrom combining herbicide treatments
Box 2. Native vegetation affected by Bitou Bush
A variety of coastal vegetation communities are affected by Bitou
Bush including sand-dune heathlands and grasslands, headland
heathlands and grasslands, various coastal woodlands, dry
sclerophyll forests, wet sclerophyll forests, and littoral rainforests. At
least 158 species are considered at risk from Bitou Bush,
including 68 species listed under the Threatened Species Act
1995 (DEC 2006). In Bundjalung National Park foredune Coast
Banksia (Banksia integrifolia) woodlands and Coastal Wattle
( Acacia sophorae) shrublands are most affected by Bitou Bush
almost completely replacing these and other species.
Coast Banksia is the dominant native species in most frontal
dune vegetation on the North Coast and an important food
source for a wide range of fauna. Flowering of Coast Banksia
occurs between late autumn and early spring in this area and
seeds are released over summer and autumn each year. Seed
dormancy may be temporarily induced by dry conditions, but
seed longevity is probably no more than 1 year. After fire, Coast
Banksia regenerates from seed or by resprouting, which varies
between habitats. The dense growth of Bitou Bush can
completely suppress regeneration of Coast Banksia leading to
long-term declines in populations of this species.
Threatened plant species most at risk of competition from
Bitou Bush in Bundjalung National Park include Chamaesyce
psammogeton, Pultenea maritima and Sophora tomentosa.
Figure 4. Bitou Bush seed is regularly dropped under CoastBanksia by birds perching in its branches. As Bitou Bush canadopt a scandent habit, it can extend into the canopy of matureCoast Banksia, effectively suppressing this shade-sensitivenative species. (Photo: T. McDonald)
emr_262.fm Page 83 Wednesday, June 28, 2006 1:22 PM
Figure 5. Extent of wildfire in Bundjalung National Park. The December 2001/January 2002 wildfire was unusually intense and burned most of thevegetation in the park, including the dune system. Although such an extensive wildfire was not ideal for a range of reasons, it provided an unprecedentedopportunity to access Bitou Bush for subsequent treatment. (Map: Department of Environment and Conservation)
emr_262.fm Page 84 Wednesday, June 28, 2006 1:22 PM
with the effects of fire (see Box 3) waspresented by an extensive wildfire thatswept through the park between 21December 2001 and 12 January 2002(Fig. 5). This wildfire occurred underextreme fire weather conditions (i.e.temperatures >40
°
C, relative humidity<10%) and was driven by strong westto north-westerly winds, severely burn-ing the southern two-thirds of Bund-jalung National Park including 258 haof Bitou Bush-infested vegetation (Fig. 3).The fire consumed Bitou Bush even onthe seaward side of the foredune, downto the vegetation limit. There was nosignificantly large area of unburntBitou Bush (Flower & Clarke 2002).Remnant natural vegetation was alsoseverely burnt with the understoreyand margins of littoral rainforest patchesaffected.
The principal control method cho-sen for post-fire treatment was aerialherbicide application, supplementedby ground-based herbicide applicationand some manual control, with appro-priate follow up for 5 years or more.Information on the impact of aerialRoundup® herbicide (2 L/ha) on nativeplants was provided by Toth (2003)who recorded 197 of 230 species ofcoastal native plants unaffected byRoundup with 25 species showingtemporary effects and eight speciesshowing some mortality. A number of
species in the treatment area, however,had never been assessed for impact.
Review of environmental factors and exclusion zones
The scale and complexity of the project,together with high biodiversity valuesand potential impact of the project,required the preparation of a detailedReview of Environmental Factors (REF).As part of the REF, a targeted florasurvey was undertaken with emphasison 101 species considered of ecologicalsignificance. This and other environ-mental considerations required theexclusion from aerial spraying of 104 haof the 510-ha project area. Areas wereexcluded because of glyphosate sensi-tivity or unknown sensitivity (littoralrainforest,
Themeda
grassland), aquatichabitats and margins, headland com-munities, threatened or significant plantlocalities, and vegetation with denseoverstorey foliage cover. The controlprogram was planned for a 5-year periodfollowed by a review to determine thesuccess or otherwise of the treatmentsand directions for further work.
Monitoring strategy
A monitoring program was establishedto examine the vegetation response tothe control program (Fig. 6). In parti-cular, the program aimed to assess the
impact of aerial spraying on bothrecently burnt and unburnt dunevegetation by addressing the followingquestions:
1
How effectively was Bitou Bushabundance reduced by aerial spraying?Was the level of control achieved inburnt and unburnt stands significantlydifferent?
2
Was there a significant impact ofaerial spraying of herbicide on therichness or abundance of nativedunal species in burnt and unburntvegetation?
The monitoring strategy involvedcollecting of data from 49 quadratsestablished in dune vegetation in April2002 – approximately 3 months beforethe first aerial spraying but 4 monthsafter the fire. This has been followed byformal observations taken twice yearlysince the project’s inception (i.e. anautumn ‘pre-spray’ observation and aspring ‘post-spray’ observation). Addi-tional quadrats were established at thethird monitoring so that approximatelyequal numbers of quadrats were moni-tored in burnt (
n
= 30) and unburntvegetation (
n
= 34).Site attributes recorded in each
quadrat included topographic position,aspect, vegetation height, vegetationlayers and the initial size distribution of
Figure 6. A schedule of progressive treatments, interspersed with monitoring events, was established early in the program and was used to guidethe sequence of works over the first 4 years, with a review to occur after 5 years. (Diagram: Andrew Benwell)
emr_262.fm Page 85 Wednesday, June 28, 2006 1:22 PM
dominant trees/shrubs. Each vascularplant species present in each quadratwas counted individually to providespecies richness and density data. Per-centage crown-cover for each species– defined as the percentage of a qua-drat covered by the horizontal projec-tion of the outer extent of live plantcrowns (Mueller-Dombois & Ellenberg1974; Walker & Hopkins 1990) – wasalso estimated visually in each 4 m
×
4 mquadrat (values of less than 1% wererecorded as 0.5%). Seedling density(numbers of seedlings of all species)in each quadrat was counted andseedling height (or lateral spread ifgreater) of selected shrub species wasestimated.
In addition to sampling burnt andunburnt vegetation, the sampling wasstratified on the basis of topographicposition (e.g. seaward and lee duneslopes), stands of Coast Banksia (
Bank-sia integrifolia
) of different height,
vegetation with greater/lesser littoralrainforest influence and incipient dunevegetation. Stratification was alsoapplied to ensure quadrats were sitedto ensure a wide geographical spreadand to ensure inclusion of some CoastBanksia to maximize the amount ofdata collected on the response of thiskey dune species.
Although the monitoring programwas initially designed to monitor theeffects of aerial spraying, about 20% ofthe plots have been subjected to othercontrol treatments as well as aerialspraying, as they have not been excludedfrom ground-spraying or manual followup applied to the general area. The datain this context therefore represent theeffects of the total control effort. Theinitial pre-fire Bitou Bush cover wasestimated for the overall site using thepercentage cover recorded from plotsin the unburnt section, prior to the firstspraying.
Post-fire vegetation condition
The 2001/2002 wildfire in the dunevegetation triggered seed germinationfrom soil and canopy seedbanks. BitouBush seedling regeneration in thesouthern burnt section outnumberedthe total native seedling population inthe first year after fire by a ratio of12:1. The mean density of native seedl-ings in post-fire regeneration was verylow, peaking at about 1.0 seedling/m
2
in June 2002, 6 months after the firebut prior to any spray treatments. Thepeak mean seedling density in ahealthy ecosystem of this type afterfire should be in the range of 50–200seedlings/m
2
(Benwell 1998).
Treatments
The first aerial spraying treatment ofapproximately 400 ha took place
Box 3. Bitou Bush – Where is its Achilles heel?
Effect of herbicideResearch carried out in the 1980s (Toth et al. 1991, 1993; Toth 2003) found that Bitou Bush was susceptible to Roundup
herbicide sprayed during winter at low doses. This application method has been found to have negligible effect on most native
species in the project area, although some are shown to the sensitive and require exclusion from spraying. As the higher
concentrations needed for summer spraying cause more damage to natives, taking advantage of this susceptibility of Bitou
Bush during winter means that aerial spraying is best to be confined to the colder weeks of the year, which on the NSW North
Coast occur mid-May to mid-July.
As flowering of Bitou Bush after post-fire germination has been observed at 8 months on the NSW North Coast, treatment
is required on an annual basis to avoid regrowth recharging soil seed banks.
Effect of fireFire is recognized as being an important factor in the ecology of Bitou Bush (Weiss 1983; CRC for Australian Weed
Management 2003). A hot fire will kill mature Bitou Bush whereas cooler fires will lead to resprouting (if the plant burns at
all). A hot fire also kills seeds in the litter and topsoil and stimulates germination of much of the remaining soil seedbank –
with the implication that if control of this initial post-fire flush occurs prior to plants maturing and seeding, then the
subsequent requirements for follow-up controls over the expected 10 years can be significantly reduced (Weiss 1983).
Conversely, if Bitou is not treated after fire, its rapid germination and resprouting has been shown to overwhelm and
outcompete natural regeneration, resulting in significant loss of biodiversity. This was seen after the 1994 fires of a similar
nature in central Yuraygir National Park, where Bitou Bush abundance increased dramatically and resources were not
available to manage the regrowth until several years later (Flower & Clarke 2002).
emr_262.fm Page 86 Wednesday, June 28, 2006 1:22 PM
over the week of 12 August 2002.Ground-based treatment occurred at thesouthern end of Shark Bay, JerusalemCreek and Snapper Rock between July2002 and January 2003. During 2003and 2004, aerial spraying of the frontaldune crest was coupled with aerialseeding of Coast Wattle (untreatedseed) and manual seeding of CoastWattle and Spinifex in selected areasaround Black Rocks.
Control programs involving bothaerial- and ground-based control haveoccurred each subsequent year, withthe amount of area controlled by bothmethods varying each year. This isdetermined by the success of previousyears’ control, extent, location and sizeof Bitou Bush regrowth, recovery ofnatural vegetation, access and availableresources. Other complementary weedcontrol has focused on scatteredoccurrences of Glory Lily (
Gloriosasuperba
) and Groundsel Bush (
Baccharishalimifolia
), a declared noxious weed.South of Shark Bay, on the Iluka Penin-sula, Bitou Bush is being progressivelycontrolled in association with extensivenatural vegetation restoration activities(Fig. 7a–c).
Results of the aerial spray treatments
Within weeks of the treatments, yellow-ing of Bitou Bush seedlings and resproutscould be readily observed and contrastedsharply with the green of the prolificpost-fire native regeneration, whichappeared unaffected by the spray (Fig. 8).Closer examination of the quadrat data,however, provided a more accuratepicture of the degree of mortality ofBitou Bush over time and the effect of theherbicide on individual native species.
Bi tou Bush control
Figure 9 shows the changes in BitouBush percentage cover in both theburnt and unburnt areas over time. Theinitial mean percentage cover of BitouBush at the end of 2001 was 55%, whichwas reduced to 5% after the third aerialspraying in 2004. Between 2004 and2005, Bitou Bush increased only slightly
from 5% to 6% cover prior to the 2005spraying. In the current (September2005) post-spray monitoring, Bitou Bushwas reduced to a mean of about 1%cover in burnt and unburnt vegetation.
The mean seedling density of BitouBush in the sampled vegetation sincethe start of monitoring is shown inTable 1. In general, Bitou Bush seed-lings have continued to germinatefrom the soil seedbank, with highernumbers of seedlings germinating inthe unburnt sites; but this is reducingover time. (Note: the higher number ofseedlings recorded in September 2005– a mean of 2 seedlings/quadrat inburnt quadrats and 11 seedlings/quadrat in unburnt quadrats (Table 1)– may be explained by the end of thedrought and the above-average rainthat occurred after the June 2005 aerialspraying.)
Changes in nat ive species cover
Figure 10 shows the changes in thecover of six frequently occurring nativespecies since the start of Bitou Bushcontrol in 2002, contrasted against thedecline in Bitou Bush cover. Four ofthe species – Tuckeroo (
Cupaniopsisanacardioides
), Prickly Couch (
Zoysiamacrantha
), Snake Vine (
Hibbertiascandens
) and Coastal Wattle – showedgradual increases in cover over time,consistent with what would beexpected during a post-fire recoveryphase. Oscillating abundance (due tospray-induced toxicity followed byrecovery) occurred for Blady Grass(
Imperata cylindrica
) and other speciesincluding Lawyer Vine (
Smilax australis
)and Hairy Clerodendron (
Clerodendronfloribundum
).Declines over time (apparently due
to spray toxicity) occurred for the grass
Paspalidium distans
, Bracken Fern(
Pteridium esculentum
), Dusky CoralPea (
Kennedia rubicunda
), Scramb-ling Lily (
Geitonoplesium cymosum
)and Native Fireweed (
Senecio lautus
ssp.
maritimus
). Other declines couldbe better explained by the life historyof individual species. For example, theshort-lived, fire ephemerals
Diplocyc-
los palmatus
and Slender Lignum(
Muehlenbeckia gracillima
) declinedover time, as would be expected overtime after fire; whereas an increasefollowed by a predictable declineoccurred for the colonizing ground-covers Blue Periwinkle (
Commelinacyanea
) and Native Violet (
Violahederacea
).It should be emphasized that these
patterns are often based on low-frequency data, meaning that althoughthe total number of samples was rea-sonably high, the majority of speciesoccurred at low frequency, less thanfive quadrats out of 66. These patternsof response are therefore tentative andbased on data trends rather than formalsignificance testing. Information on allother species is found in Benwell(2002–2005).
Changes in number of nat ive species
Although high native species coveroverall was evident on the site, datafrom the quadrats were needed toprovide more concrete informationon the species diversity (i.e. speciesrichness) of the recovering nativeplant community. Comparisons ofdata from burnt and unburnt areasenabled some insight into whichchanges may be due to the removal ofBitou Bush alone, controlling for theeffect of fire.
For example, mean species richnessin burnt quadrats has remained moreor less constant between 2002 and2005 (Table 2) as the fire’s suddenremoval of all competition is likely tohave triggered recovery of all resilientspecies present on site, with noincreases to be expected (note: thedataset analysed includes exotic andnative species). In contrast, however,mean species richness in unburnt vege-tation continued to increase gradually(Table 2) showing gradual recovery ofspecies previously suppressed by BitouBush. The increase in species richnessper quadrat in the unburnt sitesbetween April 2002 and September2005 (5.5
±
0.5 to 8.5
±
0.7) was statis-tically significant (
P
= 0.05).
emr_262.fm Page 87 Wednesday, June 28, 2006 1:22 PM
Aerial spraying of Bitou Bush costsapproximately $100/ha comparedwith the ground-based high volumespraying cost of up to $1200/ha. Manualcontrol (either by physically removingplants or cutting and painting withherbicide) can cost between $100and $500 per hectare for low infesta-tions and up to $30 000/ha for dense,multi-species infestations in areas ofhigh environmental significance(Department of Environment andConservation 2006). Voluntary inputcan assist (e.g. in a 5-ha area aroundBlack Rocks camping area the BlackRocks Dunecare group is helping toreduce Bitou Bush to a low infestationlevel and restore natural vegetation,largely by backpack spraying); butvoluntary work is not practical overlarger areas.
Figure 7. On the Iluka Peninsula, Bitou Bush is beingprogressively controlled in association with extensive vegetationrestoration activities. Iluka Bluff, for example, was heavilydominated by Bitou Bush prior to Quikspray (high volume)spraying in winter 1997 (coupled with hand work in sensitive areassuch as the remnant grassland patch). Natural regeneration wasinitially limited, due to the impacts of Bitou Bush and drought (7a)but commenced soon after soaking rains (7b), supplemented byreintroductions of species typical of the pre-existing KangarooGrass (Themeda triandra)-dominated plant community. Cover ofKangaroo Grass and other natives (7c) is now secured across thesite. (Photos: T. McDonald)
Figure 8. After herbicide was aerially applied across most of the coastal strip in winter (exceptin sensitive areas), the target species, Bitou Bush, readily died, leaving the vast majority of nativeplant species unaffected. (Photo: Jeff Thomas)
emr_262.fm Page 88 Wednesday, June 28, 2006 1:22 PM
The current integrated Bitou Bushcontrol program has an annual averageexpenditure of approximately $75 000.This can be broken down into $40 000
for contract aerial and ground-basedspraying, $15 000 for contract moni-toring and $20 000 for in-kind ground-based control and project management.
Implications for management
Has Bi tou Bush control been ef fect ive to date?
It appears the program has beensuccessful to date in reducing BitouBush cover, with 4 years sustainedcontrol reducing its cover from 55%pre-fire to 1%, 3 months after the winter2005 control. Results from the last 2years show that cover of Bitou Bush isreturning to an average of 5–6% justprior to control. This number reflects apattern in the landscape where themajority of the project sites have verylow densities of Bitou Bush but thereare still significant areas with higherdensities of Bitou Bush. Sites with higherdensities result from previously highercover of Bitou Bush (and lower levelsof native species resilience – see Fig. 11)combined with less effective controlin some cases, for a variety of reasons.Aerial spraying in 2002 and 2003 wasunsuccessful in achieving the requiredkill rate in a large part of the projectsite, resulting in plants maturing andflowering and seeding, thereby re-plenishing the soil seed bank.
Achieving uniform control annuallyover a large area has proven to beimpractical, with plants missed due tosheltering effects from canopy treesand the difficulty in treating Bitou Bushin some exclusion zones. In the firstyear, no exclusion zones around water-bodies were treated because of lack ofaccess and some dune crests wereexcluded because of uncertainty aboutpotential dune erosion (Fig. 12). Thedevelopment of aerial spot sprayingtechnology and adaption of newerboom spraying technology in the lastfew years has improved the ability tocontrol more remote and inaccessible
Table 1. Bitou Bush mean seedling density (seedlings per 4 m2) ± standard error in burnt and unburnt dune vegetation during Bitou Bush sprayingat Bundjalung National Park
Figure 9. Change in the mean percentage cover of Bitou Bush in burnt and unburnt quadrats.Falls in cover were recorded post-spray (in spring) and increases were recorded pre-spray (lateautumn), except for the first fall in burnt quadrats, which was of course due to fire.
Figure 10. Changes in the mean percentage cover of Bitou Bush and six of the most frequentlyoccurring native species in 4 years averaged over all quadrats. The four decreases in Bitou Bushoccurred after aerial spraying each year. Blady Grass has a similar oscillating pattern due to die-offof foliage after spraying then regrowth from rhizomes. Other native species show increasing cover-abundance, particularly Coast Wattle.
emr_262.fm Page 89 Wednesday, June 28, 2006 1:22 PM
infestations and infestations in environ-mentally sensitive areas.
The reduction in Bitou bush germi-nation rate over time as shown inTable 1 is another indication that thetreatment program is achieving timelycontrol, although the relatively highernumbers in the unburnt area point tothe need to maintain a similar level ofcontrol for at least few more years.
How long wi l l i t take?
The time required for completion ofthe project as a whole is yet to bedetermined and will depend to a largeextent on the level of control requiredto achieve the initial aim of biodiversityprotection with significant Bitou Bushreductions and recovery of native
vegetation. Little published data existon the timeframe required to achieve aspecified level of control. We do know,however, that discontinuation of aproject too early, or reduction ofresources below the rate required tomaintain control, can result in reinfes-tations. In addition, it is known thatinfestations on the North Coast willreturn to pre-existing densities within2 years after a single control unlessfollow up is implemented (Thomas,pers. obs., 2002) and experience hasshown that ongoing maintenance forminimum of 5 and more likely 10 yearsis required to achieve sustained re-ductions in infestations
The change in Bitou Bush infestationsvaries with location and is influenced
by the existing soil seedbank and seeddispersed into the treated areas fromnearby untreated infestations. Compe-tition from existing native vegetationand native species germinating fromsoil-stored seed and suppressed root-stocks (i.e. ecosystem resilience) alsoaffects Bitou Bush densities. In areaswith low resilience (such as some pre-viously sandmined but unrevegetatedareas), supplementary revegetation withnative species is required to providecompetition for Bitou Bush.
The control program has nowreached a stage where after this year’streatment the annual broadacre aerialspraying is likely to be no longerrequired in terms of achieving cost-effective control of large infestations,
Table 2. Mean species richness ± standard error in burnt and unburnt dune vegetation during Bitou Bush spraying at Bundjalung National Park
Figure 11. Condition mapping carried out at Black Rocks visitor area shows that, prior to the wildfire and subsequent herbicide treatment in 2000,Bitou Bush was dominant (Classes 3–5) along the frontal dunes and particularly high where associated with unstable dunes. Reassessment in 2005,5 years after aerial and ground-based herbicide treatments (carried out annually in winter), found Bitou Bush rarely occurring in most of the site, althoughclusters persisted in some areas where treatment was either inconsistent or the original infestation was most intense. (Adapted from: McDonald 2002)
emr_262.fm Page 90 Wednesday, June 28, 2006 1:22 PM
which has to be balanced against theimpacts of spraying native species.Current and further stages of the pro-gram are therefore likely to combinethe newer targeted aerial applicationmethods with increased on-groundeffort, focusing on exclusion zones, areasof persistently high Bitou Bush coverand existing or potential environmen-tal impact. Work will also continue onthe timely control of other weedspecies such as Glory Lily (Gloriosasuperba), Lantana (Lantana camara)and Winter Senna (Senna pendula var.glabrata) which can proliferate afterBitou Bush control.
Impact of aer ia l spraying on nat ive vegetat ion
Aerial spraying elsewhere and on thissite has been found to have negativeimpacts on a range of native species.The decision to use the treatment atthis site, however, was influenced byweighing up these costs against theanticipated economic and ecologicalbenefits of Bitou removal using thismethod in the preliminary phase, priorto more use of better-targeted methods.
The benefits of this decision appearto have been borne out by the resultsto date. Seedlings of Coast Wattle
(Acacia sophorae) and Coast Banksia(Banksia integrifolia), the two domi-nant native species in the coastal duneecosystem, appeared to be unaffectedby aerial spraying. Most other nativespecies showed no observable affectswith effects of spray observed only forthe species listed in Table 3.
Further observat ions on the recovery of the plant communit ies
Seedling densities throughout the sitewere initially lower after fire thanwould be expected at a healthy site –possibly due to the effect of Bitou Bushover recent decades.
During 3 years of monitoring, littoralrainforest species exhibited little tono seedling recruitment. Sclerophylltrees, shrubs and herbaceous speciesproduced more seedlings than littoralrainforest species, but mean densitieswere still very low.
Monitoring did not detect theexpected seedling numbers or res-prouting of Coast Banksia immediatelyafter the 2001/2002 fire; both in thetaller woodland ecotype and in muchof the stunted shrubland ecotype. Thepoor initial regeneration is likely tohave been due to very low seedlingrecruitment in the Coast Banksia wood-land ecotype and failure of resproutingin the Coast Banksia shrubland ecotype.In the taller ecotype, competition withBitou Bush over a decade or more mayhave depressed flowering and seedproduction and the post-fire recoveryof Bitou may also have smotheredseedlings. Another factor, however,may have been that the fire occurred atthe point of seed shed, reducing regen-eration potential. In the stunted ecotype,Bitou Bush competition before fire (or
Table 3. Species that exhibited post-spray foliage dieback and decrease in mean crown-cover and/or seedling death in the study area (Benwell 2002–2005)* indicates exotic species
Figure 12. Dune crests at Black Rocks were among areas excluded from aerial spraying in thefirst year of treatment due to concern about potential for erosion. This delay (coupled with the readyresprouting of layering Bitou on dune crests) has meant that control in these sites has been moregradual than other areas. This gradual treatment – coupled with direct seeding of sites at BlackRocks with Coastal Acacia and Spinifex – has resulted in secure revegetation and stability of dunecrests at Black Rocks. (Photo: T. McDonald)
emr_262.fm Page 91 Wednesday, June 28, 2006 1:22 PM
sudden exposure to salt-laden winds)may have reduced the vigour of CoastBanksia, as several individuals producedinitial suckers then died in the firstyear after fire.
Despite this, Coast Banksia stillexists over much of the southern burntsection in 2005 (3 years after fire) and,has recruited well over much of thenorthern burnt section, possibly sinceeffective control of the highly compet-itive Bitou Bush. The surviving CoastBanksia population should continue toincrease in number and vigour withthe ongoing control of Bitou Bush.
Concluding commentsIn summary, this project has shownthat Bitou Bush control using a range oftreatments including extensive aerialapplication of low concentrations ofherbicide has been successful both inburnt and unburnt areas. After fire, burntsites have demonstrated a more rapidreduction of Bitou Bush germinationthan have unburnt sites. Although thereare clear benefits of aerial spraying,caution is required to ensure sensitiveplants and plant communities are notadversely affected and that adequatefollow up – by increasingly moretargeted methods – is available toprevent the recovery of Bitou Bush andthe spread of other weeds that maybenefit from the reduction of thishighly competitive weed species.
at the Department of Environment andConservation (DEC), for the provisionof maps. We also thank other DEC staff,contractors and volunteers who havecontributed to this control program.
References
ARMCANZ, ANZECC and ANZFM (2000) Weedsof National Significance Bitou Bush and Boneseed(Chrysanthemoides monilifera ssp. rotundataand monilifera ) Strategic Plan. Agriculture &Resource Management Council of Australia &New Zealand, Australian & New Zealand Envi-ronment & Conservation Council and ForestryMinisters. National Weeds Strategy ExecutiveCommittee, Launceston.
Benwell A. S. (1998) Post-fire seedling recruitmentin coastal heathland in relation to regenerationstrategy and habitat. Australian Journal of Botany46, 75–101.
Benwell A. S. (2002a, 2002b, 2003a, 2003b, 2004,2005) Monitoring of Plant Species Composi-tion in Recently Burnt and Unburnt FrontalDune Vegetation Subject to Aerial Spraying ofBitou Bush (Chrysanthemoides moniliferasubsp. rotundata) in Bundjalung National Park.Reports to the Department of Environment andConservation (NSW), Hurstville, NSW.
CRC for Australian Weed Management (2003)Weed Management Guide: Bitou Bush. CRCfor Australian Weed Management, Adelaide, SA.
Department of Environment and Conservation(NSW) (2006) Threat Abatement Plan for Inva-sion of Native Plant Communities by Chrysan-themoides monilifera (Bitou bush and boneseed).Department of Environment and Conservation(NSW), Hurstville, NSW.
Flower P. and Clarke M. (2002) The Control ofBitou Bush by Aerial Application of Glypho-sate, Bundjalung National Park 2002–06, Vol. 1Environmental Review. Internal report to theNSW National Parks and Wildlife Service,Grafton, NSW.
Griffith S. J. (1983) Vegetation Survey of Bund-jalung National Park. Internal report to theNSW National Parks and Wildlife Service.Hurstville, NSW.
Harden G. J. (ed.) (1990–93) Flora of New SouthWales, Vols 1–4. New South Wales UniversityPress, Kensington, NSW.
McDonald T. (2002) Bundjalung National Park Veg-etation Restoration and Rehabilitation Plan(Incorporating Pest plant survey and controlguidelines). National Parks & Wildlife Service,Alstonville, NSW.
Mueller-Dombois D. and Ellenberg H. (1974) Aimsand Methods of Vegetation Ecology. JohnWiley and Sons, New York.
Natural Heritage Trust (2003) The Work Doesn’tStop on Award-Winning Iluka Bluff Beach.Natural Heritage Trust, No. 16, p. 11. Departmentof Environment and Heritage, Canberra, ACT.
NSW National Parks and Wildlife Service (2001)NSW Bitou Bush Strategy. NSW NationalParks and Wildlife Service, Hurstville, NSW.
Scanlon T. (2001) NSW North Coast Bitou BushManagement Strategies. North Coast WeedsAdvisory Committee.
Toth J. (2003) Toxicity of Roundup® to Native andSome Introduced Plants. Management ofEnvironmental Weeds Pty Ltd, Mona Vale, NSW.
Toth J., Milham P. J. and Maguire M. J. (1991)Seasonal differences in tolerance of glyphosateand metsulfuron methyl by Bitou bush and fourplants species indigenous to coastal NSW.Proceedings of the 6th Noxious Plants Confer-ence 4, 18–20.
Toth J., Milham P. and Nazer C. (1993) Aeriallyapplied herbicides selectively control Bitou bushgrowing in communities of indigenous plants onsand dunes. Proceedings of the 14th AsianPacific Weeds Conference, Vol. 2, pp. 67–71.Asian-Pacific Weed Science Society.
Toth J., Milham P. J., Meszaros I., Kaldor C. J.,Fullerton R. N. and Burrows F. (1996) Researchon chemical control of Bitou bush in New SouthWales. Proceedings 11th Australian WeedsConference, Melbourne, pp. 468–475. WeedScience Society of Victoria, Frankston, Victoria.
Walker J. and Hopkins M. S. (1990) Vegetation. In:Australian Soil and Land Survey Field Handbook,2nd edn. (eds R. F. McDonald, R. F. Isbell,J. G. Speight, J. Walker and M. S. Hopkins),pp. 58–86. Inkata Press, Melbourne.
Weiss P. W. (1983) Invasion of Coastal AcaciaCommunities by Chrysanthemoides. PhDThesis, Australian National University, Canberra,ACT.
Weiss P. W., Adair R. J. and Edwards P. B. (1998)Chrysanthemoides monilifera (L.) T. Norl. In:The Biology of Australian Weeds (eds F. D.Panetta, R. H. Groves and R. C. H. Shepherd),Vol. 2, pp. 49–61. R.G. and F.J Richardson,Melbourne, Vic.
emr_262.fm Page 92 Wednesday, June 28, 2006 1:22 PM
