20 Protection Quarterly Vol.24(1) 2009 lack of suitable environmental conditions for germination and establishment of na- Ten year post-fi re response of a native ecosystem in tive species. Turner et al. (in press) also the presence of high or low densities of the invasive suggested that recovery of invaded sites may not occur after bridal creeper control. weed, asparagoides They established that the readily germi- nable seed bank below bridal creeper in Peter J. TurnerA,B,C,E and John G. VirtueB,D south west Australia contained very few A School of Animal Biology, University of Western Australia, Stirling Highway, native species compared to exotic spe- Crawley, Western Australia 6009, Australia. cies. Further studies were recommended B CRC for Australian Weed Management. to investigate the role of fi re as a restora- C tion tool following bridal creeper control; CSIRO, Division of Entomology. given that fi re has an important role in D Department of Water, Land and Biodiversity Conservation, GPO Box 2834, the regeneration of some Australian na- Adelaide, South Australia 5001, Australia. tive communities (Shea et al. 1979, Bell et E Present address: Pest Management Unit, Department of Environment and al. 1993, Fisher 1999). For example, seeds Climate Change, PO Box 1967, Hurstville, New South Wales 1481, Australia. of a large number of native leguminous shrubs, such as acacias, require heat before germination occurs (Shea et al. 1979, Bell et al. 1993, McCaw 1998) and these spe- Summary cies would not have been recorded in the Bridal creeper, the regeneration of some native Austral- readily germinable seed bank measured (L.) Druce, is a major environmental ian and speed up the recovery by Turner et al. (in press). The use of fi re weed in southern Australia. Being a geo- of bridal creeper-invaded ecosystems, as a restoration tool could help increase phyte, it has annual shoot growth with provided that bridal creeper and other the germination rate of native species and a large tuberous root mat belowground. weeds are kept at a low post-fi re density, may help tip the balance back towards na- It is capable of displacing native vegeta- naturally or through targeted control. tive species, by increasing the ratio of na- tion and has been targeted for control Key words: bridal creeper, environ- tive to exotic germinations (Turner et al. in Australia, especially using biological mental weeds, soil nutrients, succession, in press). control. Previous studies on the impact of weed impacts. Turner and Virtue (2006) also observed this environmental weed have suggested that dead bridal creeper tubers remained that without further restoration, invaded Introduction in the experimental plots eight years after areas could take many years to recover. In Australia, bridal creeper is a Weed of the weed had been killed with herbicide, As fi re can be used as a restoration tool, National Signifi cance (WoNS) (Thorp and and this may have also impacted on na- given it can stimulate the regeneration Lynch 2000). A garden escapee, this geo- tive species recovery. Turner et al. (2006) of some native Australian plants, this phyte is able to invade undisturbed na- established that this belowground bio- study aimed to determine the response of tive ecosystems (Hobbs 1991, Raymond mass, with its slow decomposition, may a native plant community for the fi rst ten 1995) and impacts or threatens many na- pose a problem for many years after con- years following a fi re, with and without tive plant species (see Fox 1984, Sorensen trol. As a way to deplete this root system, the presence of bridal creeper. and Jusaitis 1995, Willis et al. 2003, Wil- Willis et al. (2003) recommended that fi re Following a wildfi re in March 1996, lis et al. 2004, Downey 2006). For exam- could be used to destroy the new season’s plots were established in a mallee rem- ple, bridal creeper is the primary threat foliage in autumn, followed by treating nant in South Australia. In October 1996, to the largest remaining population of the re-growth with herbicide. However, bridal creeper was controlled in half the the endangered shrub, spicata Carr (1996) suggested that fi re was not an plots using glyphosate. In 2006, there was R.Br. in the south-west of Sydney (Willis appropriate management tool for bridal still a signifi cant difference in the density et al. 2003). Community impacts are also creeper as it stimulated bridal creeper of bridal creeper, with 33.4 ± 5.0 (mean ± evident, with plant assemblages in areas growth and invasion. Fire has also been SE) emerging shoots m−2 in the untreated invaded by bridal creeper in south west reported to promote the spread and den- plots compared to 9.1 ± 1.2 shoots m−2 in Australia containing 52% fewer native sity of a number of other exotic plant spe- the controlled plots. However at the same plant species compared to nearby native cies (Adams and Simmons 1991, Milberg time, there was no signifi cant difference reference areas, which contained little or and Lamont 1995, Briese 1996, Duggin and in the native plant assemblages or with no bridal creeper (Turner et al. in press). In Gentle 1998). In addition, when smoke the number of native plant species be- an eight year removal experiment, Turner water was used to promote the germina- tween plots with high or low density of and Virtue (2006) established that bridal tion of the seed bank collected from bridal bridal creeper, except for small shrubs, creeper could reduce biomass and abun- creeper sites in the study by Turner et al. creepers and climbers which had higher dance of native plants. However within (in press), it assisted in the germination of cover in the untreated plots. A difference this eight year timeframe, there was no both native and exotic species. Therefore, in soil nutrients was also evident. The signifi cant change detected in native plant the use of fi re in bridal creeper manage- soil where bridal creeper was not control- species richness following bridal creeper ment needed to be investigated given that led had signifi cantly higher ammonium, control and only one saltbush, Enchylaena it could promote further weed invasions. potassium and organic carbon, compared tomentosa R.Br., and native grasses showed The biology of bridal creeper has recent- to where bridal creeper was controlled. signifi cant increases in cover. ly been described (see Morin et al. 2006a). This study site, in the early stages of post- Luken (1997) recommended that man- In its native range in southern Africa, it fi re succession, appears to be resilient to agement activities aimed at removing mainly occurs as an understorey species the impacts of bridal creeper and other exotic plant species should involve the and is usually found scrambling or climb- weed species, with acacias and other na- manipulation of both exotic and native ing up other plants (Kleinjan and Edwards tive trees and shrubs dominating the site. species. Turner and Virtue (2006) sug- 1999). Within Australia, bridal creeper has It is therefore concluded that fi re can be gested that the minimal recovery of na- the potential to dominate native vegeta- an important restoration tool to stimulate tive species in their study could relate to a tion both above and belowground, with Plant Protection Quarterly Vol.24(1) 2009 21 climbing annual shoots and extensive sampled within the plots. In August 1996, 2001). ANOSIM (Analysis of Similarity) belowground storage tubers (Raymond the number of shoots of bridal creeper was was used to determine differences in com- 1996). However, simply reducing the pres- recorded. This was repeated in November munity composition between the removal ence of bridal creeper may not guarantee 1998 and May 2006. From April 1997, as na- and bridal creeper plots within the burnt successful restoration of invaded areas tive plants began to recover (re-sprouting area. and additional restoration efforts may be and germinating) after the fi re, the pres- needed to ensure the protection of native ence of all other plant species within the Results vegetation (Turner et al. in press). plots was recorded as well as the percent- Before treatments were applied, the This paper reports on a complementary age aerial shoot cover of each species. This number of bridal creeper shoots was not study to that of Turner and Virtue (2006) was repeated in May 2002 and May 2006. different between bridal creeper plots and which determined the impact by bridal At the end of the experiment in May 2006, removal plots in the burnt area (F = 0.75; creeper, by measuring the response of the biological control agent, the bridal d.f. 1,28; P = 0.395). In August 1996, the vegetation following the removal of brid- creeper rust ( myrsiphylli (Thuem.) bridal creeper (untreated) plots had 29.7 al creeper. Similarly, the aim of this study Wint.) was observed to be at an early stage ± 3.9 shoots m−2 (mean ± SE), while the was to measure the impact by bridal creep- of establishment at the study site. removal plots had 25.5 ± 3.0 shoots m−2 er on plant species and to measure their Soil measurements were also taken in prior to being treated with herbicide (Fig- response to bridal creeper removal. How- May 2006. From the centre of each of the 30 ure 1). Following the treatment and ten ever unlike Turner and Virtue (2006), this plots, soil cores were taken from below the years later in 2006, there was a signifi cant study was conducted following a wildfi re. bridal creeper tuber mats. Soil cores 5 cm difference in the density of bridal creeper Therefore, the study reported here aimed deep and 5 cm in diameter were taken 5 (F = 26.28; d.f. 1,28; P <0.001), with 33.4 to investigate the successionary pathway cm below the litter layer. Cores from three ± 5.0 emerging shoots m−2 in the bridal that occurred over the fi rst ten years post- proximal plots of the same treatment were creeper (untreated) plots compared to 9.1 fi re within a mallee community, with or bulked and any large organic matter was ± 1.2 shoots m−2 in the removal (herbicide without the presence of bridal creeper. As removed, such as tubers and other roots. treated) plots. The maximum number of nutrient availability in the soil increases Five samples per treatment were then bridal creeper shoots in a bridal creeper following fi re (Hobbs and Huenneke 1992, forwarded in airtight containers to CSBP plot was 77.9 ± 12.4 shoots m−2. Handreck 1997), this study also measured Laboratories (WA) for chemical analysis. Six months after the herbicide treat- − + the impact of bridal creeper and of the fi re Nitrogen (NO3 and NH4 ), extractable ment was applied, in April 1997, the mean on the soil nutrient status. phosphorus, potassium, organic carbon cover of native plant species was similar and pH were measured. between bridal creeper plots and removal Materials and methods Also in May 2006, the soil sampling plots (Figure 2). In May 2006, there was Study area was repeated in a third treatment, an un- no signifi cant difference in the number of The study was undertaken in Meningie burnt area adjacent to the burnt plots that native plant species between treatments Hill Reserve, a mallee remnant commu- had a similar pre-fi re coverage of bridal (Table 1). There were also no signifi cant nity, south-east of Adelaide near the town- creeper. Five additional soil samples were differences in cover of individual groups ship of Meningie, South Australia 35°41’S, taken from this unburnt area. Ten 1 m × 1 of native or exotic plant species, except for 139°21’E. The study site has calcareous, m quadrats were selected in the unburnt native small shrubs and native climbers sandy soil with a south-westerly aspect. section of the reserve, based on an even and creepers, where their cover was high- Native vegetation within the reserve is and dense cover of bridal creeper. The er in the bridal creeper (untreated) plots dominated by an overstorey of Eucalyptus number of shoots of bridal creeper was (Table 1 and Figure 2). In contrast, the year diversifolia Bonpl. and E. incrassata Labill., recorded and two soil cores were taken immediately following the application of with large shrubs being mainly Acacia spp. from each quadrat and bulked. Soil sam- glyphosate (1997) there was no signifi cant and Melaleuca lanceolata Otto. In March ples from two quadrats were further difference in the cover of these native small 1996, a wildfi re occurred in part of the re- bulked and a total of fi ve samples from shrubs (F = 0.17; d.f. 1,28; P = 0.684, Figure serve, burning most of the aboveground this unburnt area were also forwarded to 2) and the cover of these native climbers vegetation (including bridal creeper) ex- CSBP Laboratories for analysis. Except for and creepers (log10 transformed; F = 0.45; cept for woody trunks of the mallee euca- the number of shoots of bridal creeper, no d.f. 1,28; P = 0.509, Figure 2) between brid- lypts. other vegetation was measured within this al creeper plots and the removal plots. In Methods followed here were similar to unburnt area. 1997, native small shrubs averaged 1.02% Turner and Virtue (2006), who undertook Statistical analysis was carried out us- ± 0.30 in the bridal creeper plots and 0.84% an eight year bridal creeper removal study ing GenStat (2003). Transformations were ± 0.30 in the treated removal plots, while near Owen, north of Adelaide. In August applied when appropriate, via a log10 native climbers and creepers averaged 1996 at Meningie Hill, 30 (3 m × 3 m) plots transformation, to meet the assumption 2.47% ± 1.36 in the bridal creeper plots and were chosen where the fi re had burnt fi ve of homogeneity of variances. If signifi cant 1.34% ± 0.41 in the removal plots. months previously and where there was differences were found in the above soil In May 2006, there was no signifi cant also an even coverage of bridal creeper. analyses, it was followed by a pair wise difference in the native plant assemblages Bridal creeper had re-sprouted from the Tukey comparison test (GenStat 2003). between treatments (ANOSIM R = 0.017; belowground root system following the Multivariate analysis was also used to P = 0.274, Figure 3). Acacias and large fi re. Bridal creeper was then controlled in compare the plant assemblages sampled shrubs and trees dominated both treat- half the plots in October 1996, using 33% in May 2006, to compare differences be- ments in 2006. Combined cover of these Roundup® (360 g L−1 glyphosate) with 2% tween removal (herbicide treated in the large woody native plants within the Pulse Penetrant® (1020 g L−1 polyether burnt area) plots and bridal creeper plots bridal creeper plots averaged 46.5%, while modifi ed polysiloxane), applied by hand (untreated in the burnt area) (Primer 6, in the removal plots this averaged 55.0%; with a sponge to minimize off-target con- Clarke and Warwick 2001). A Bray-Cur- yet there was no signifi cant difference be- tact with native plants. These 15 plots tis similarity index on square-root trans- tween treatments (Table 1). were labelled as ‘removal plots’ and the formed percentage cover of all plant spe- Even though no differences were de- untreated 15 plots were labelled as ‘bridal cies (excluding bridal creeper) was calcu- tected in the native plant assemblages creeper plots’. Between August 1996 and lated to construct a rank-similarity matrix (Figure 3), the soil in the burnt area be- May 2006, vegetation was periodically (see Clarke 1993, Clarke and Warwick neath the untreated plots (bridal creeper 22 Plant Protection Quarterly Vol.24(1) 2009 50 Discussion Bridal creeper plots Removal plots Originally the aim of this experiment was to investigate the successionary path- 40 way that occurred ten years post-fi re in a mallee community, with or without 30 the presence of bridal creeper. However, (±SE) bridal creeper had begun to recover in −2 the removal (herbicide treated) plots, to 20 a density of 9.1 ± 1.2 shoots m−2 in 2006. This was still signifi cantly lower than in shoots m the untreated plots, therefore this study

Mean no. of bridal creeper Mean no. 10 measured the response of the plant com- munity following a fi re in the presence of high and low densities of bridal creeper. 0 However, neither density of this invasive 1996 1998 2006 weed infl uenced the post-fi re response of Year vegetation within this mallee ecosystem, with acacias and other native trees and Figure 1. The abundance of bridal creeper across the period of the study. shrubs dominating the burnt area of site The bridal creeper plots (not treated with herbicide) and the removal plots 10 years post-fi re. (herbicide treated) were all located within the burnt area of the study site. An invader may have a large effect The removal plots were treated with glyphosate two months following in one habitat, but a negligible effect in the 1996 measurements. Bridal creeper abundance was not recorded in the another habitat (Byers et al. 2002). In por- tions of the adjacent unburnt area, bridal years not shown. In 2006, there was a signifi cant difference in the density of creeper density was relatively high, simi- bridal creeper between bridal creeper plots and removal plots (F = 26.28; d.f. lar to that observed at the Owen site in 1,28; P <0.001). the study by Turner and Virtue (2006), which had an average of 87.8 ± 8.8 shoots m-2 in 2004 (mean ± SE, Turner and Virtue Table 1. Differences in number of native species per plot (mean ± SE) and unpublished). However, whilst not di- A plant cover (%) between treatments in May 2006 . rectly measured, individual shoots were Variable Bridal creeper Removal FPshorter throughout the whole Meningie (untreated) plots (herbicide site compared to Owen, with the former treated) plots site appearing to have less bridal creeper aboveground biomass and rarely having No. of native species 11.20 ± 0.62 10.33 ± 0.41 1.36 0.253 tall ‘curtains’ of bridal creeper shoots. The Cover of: whole study site at Meningie may be less Native small shrubsC 3.73 ± 1.08 1.68 ± 0.61 5.46 0.027B suitable for bridal creeper compared to the Owen site. Native creepers and 3.16 ± 0.47 1.82 ± 0.28 7.16 0.012B Hester and Hobbs (1992) established a climbers strong correlation between soil phospho- Native monocots 3.65 ± 1.70 2.95 ± 1.50 0.19 0.668 rus and abundance of exotic species. The Native acacias 19.75 ± 4.59 19.89 ± 5.79 0.67 0.419 Owen site had available soil phosphorus in the range of 10.8–13.8 mg Kg−1 (Turner Other native trees and 26.70 ± 2.97 35.14 ± 4.03 2.84 0.103 and Virtue 2006), compared to 3.6–6.0 mg large shrubs Kg−1 at the Meningie site (Table 2). Lower Exotic grasses and 3.45 ± 1.20 1.45 ± 0.27 3.06 0.091 soil phosphorus, combined with reduced bulbsD aboveground biomass, may partly ex- A All plots were within the burnt area. The analysis of variance models with treatment plain the lack of competitive effects of bridal creeper observed in this post-fi re (n = 15) as the only factor. A log10 transformation was applied to all variables, except no. of native species and native trees and large shrubs before the analysis. experiment. However, in some parts of the B Statistically signifi cant (P < 0.05). unburnt area at this Meningie study site, C See Figure 2 for description of native groups. there were low levels of phosphorus (Ta- D Exotic grasses and bulbs were dominated by veldt grass (Ehrharta calycina Sm.) and ble 2) and bridal creeper abundance was Freesia sp. relatively high, averaging 94.0 ± 9.4 shoots m−2. Therefore, at this Meningie study site, bridal creeper was still able to achieve a high abundance in some of the unburnt plots) had signifi cantly more ammonium, the burnt bridal creeper and unburnt plots area that had relatively low soil nutrients. potassium and organic carbon compared having approximately eight and twenty Therefore, it is likely that the native veg- to the burnt, removal (herbicide treated) times the number of shoots found in the etation, being the primary post-fi re succes- plots, ten years after the fi re (Table 2). The burnt removal plots respectively (Table 2). sional species such as acacias, were able to additional plots sampled in the unburnt In the burnt area, the bridal creeper centre at least partially suppress bridal creeper in area, which contained bridal creeper, had sub-plot (1 m × 1 m) had 38.2 ± 7.2 shoots the areas that were burnt. a signifi cantly higher pH but lower organ- m−2 (mean ± SE); the removal centre sub- It appears, from the several studies that ic carbon and ammonium in comparison plot had 4.8 ± 1.0 shoots m−2; while in the have investigated the response of plant to the burnt bridal creeper plots. The ar- unburnt area, the ten additional quadrats communities to bridal creeper control, that eas where the soil was sampled had dif- averaged 94.0 ± 9.4 shoots m−2. impacts and the recovery of native vegeta- ferent abundances of bridal creeper, with tion, as well as the response of secondary Plant Protection Quarterly Vol.24(1) 2009 23 invaders, are site specific and depend In the burnt area of this Meningie study, An alternative hypothesis for the high- on the history of the site and its current native small shrubs, creepers and climbers er cover of native small shrubs, creepers successionary state. At the Owen site in had higher cover in the bridal creeper plots and climbers in the bridal creeper plots South Australia, Turner and Virtue (2006) compared to the removal (herbicide treat- compared to the removal (herbicide treat- reported an increase in biomass of cheno- ed) plots (Table 1). These species could be ed) plots at this Meningie site is that they pods but overall there was no difference responding to the higher nutrients found may have been accidentally killed by the in the number of species following bridal below bridal creeper. In Western Australia, herbicide application in the removal plots creeper control. There was also an increase at the same study site mentioned above, during the initial bridal creeper removal in in the exotic Oxalis pes-caprae L. eight years Quell Creek, Turner et al. (2008) also re- 1996. Whilst due care was taken to only tar- after this control. Similar to this Meningie ported that native climbers had increased get bridal creeper, there may have been in- study, four sites in a southern Western Aus- from 0.07% to 5.0% and that this cover of advertent contact with some regenerating tralia study showed no overall signifi cant native climbers was higher than the nearby native species. However, it is assumed that increase in exotic cover following bridal native reference plots at the same time in the off-target impacts were minimal given creeper control (Turner et al. 2008). In the 2006, being only 0.8% (Turner et al. 2008). that there was no signifi cant difference in Western Australia study, in bridal creeper In this Western Australia study at Quell the cover of native small shrubs, creepers invaded plots, cover of other exotic species Creek, the bridal creeper invaded plots and climbers in 1997, a year following the (excluding bridal creeper cover) increased had higher soil nutrients than the nearby application of glyphosate. Even so, the dif- from 4.2% in 2004 to 12.8% in 2006 follow- native reference plots (Turner et al. submit- ferences in cover of these native species in ing the biological control of bridal creeper; ted) and, following biological control, na- 2006 may not be ecologically signifi cant however this was comparable to the exotic tive climbers had six times greater cover in (although statistically signifi cant between cover in the nearby uninvaded reference the higher nutrient soil (Turner et al. 2008). treatments), given the differences recorded plots at the same time in 2006 (Turner et This is also supported by a glasshouse trial relate to a difference in cover of only one al. 2008). Yet, at one of the Western Aus- that established that a native climber, Bil- to two percent between treatments and be- tralia sites, Quell Creek, Turner et al. (2008) lardiera heterophylla (Labill.) L.W.Cayzer & tween 1997 and 2006 (Figure 2). reported an increase in exotic cover from Crisp) (bluebell creeper), had increased Nutrient availability in the soil is usual- 0.03% to 23.4% following the biological growth rates with increases in soil fertility ly only increased for a short time following control of bridal creeper. (Turner et al. submitted). fi re (Hobbs and Huenneke 1992, Handreck

40 Small shrubs Creepers and climbers 35 Native monocots Acacias 30 Native trees

25

20

15

10 Mean percentage cover (± SE) Mean percentage cover

5

0 1997 2002 2006 1997 2002 2006 Bridal creeper plots Removal plots

Figure 2. Percentage cover of native species in 1997, just after the herbicide treatment was applied, and in 2002 and 2006. Both bridal creeper and removal plots were burnt. Bridal creeper plots were not treated with herbicide. Native trees are a combination of Eucalyptus diversifolia Bonpl., E. incrassata Labill. and Melaleuca lanceolata Otto. Acacias are a combination of Acacia paradoxa DC. (primarily), A. pycnantha Benth. and A. longifolia (Andrews) Willd. Native monocots are mainly a combination of Dianella revoluta R.Br., Xanthorrhoea caespitosa D.J.Bedford and orchids. Creepers and climbers were dominated by Billardiera cymosa F.Muell. and Carpobrotus modestus S.T.Blake. Small shrubs were dominated by Thomasia petalocalyx F.Muell., with fewer Rhagodia sp., Dampiera sp. and Hibbertia sp. 24 Plant Protection Quarterly Vol.24(1) 2009 1997). Bridal creeper may be able to per- stimulated bridal creeper growth and in- It seems plausible that the native veg- manently capture and retain the nutrients vasion. Given this, fi re should not be used etation was able to suppress bridal creep- released during the fi re by trapping nutri- as a sole control measure for bridal creep- er and other weeds in the burnt area at ents and organic material with its dense er, as bridal creeper will re-shoot and trap Meningie, for the ten years following the tuberous root mat (Turner et al. 2006) and the nutrients released during the fi re. This single disturbance event, being a wildfi re. recycling them through its annual senes- could over time transform the community However, in the absence of future fi res, cence. This may explain the differences in and move it further away from its restora- it has been reported that populations of nutrient levels between the burnt bridal tion goals. Therefore, fi re is only suitable if native legume species (including acacias) creeper plots and the burnt removal plots used in conjunction with other restoration decline, as adults die at the end of their in this study (Table 2). Carr (1996) sug- and control techniques, such as those rec- normal life span of approximately four gested that fi re was not an appropriate ommended by Willis et al. (2003). to 15 years without seedling recruitment management tool for bridal creeper as it (Shea et al. 1979, McCaw 1998, Fogarty and Facelli 1999). Therefore this site may become more vulnerable to invasion at a later successional stage. The burnt areas where bridal creeper was not controlled could be more vulnerable to invasion by other weeds or to an increase in the den- sity of bridal creeper, given the differences May 2006 observed in the soil nutrients (for example see Lake and Leishman 2004 and Turner Bridal creeper et al. submitted). Following the decline of Removal early post-fi re species, such as acacias, ar- eas with the higher nutrient levels may be replaced with those species (native and exotic) that respond more favourably to higher nutrient soils. Biological control agents have been released in Australia to control bridal creeper (Morin et al. 2006b, Turner et al. 2008). Briese (1996) suggested that fi re and biological control may not be com- patible strategies for weed management in native areas, although a personal ob- servation in a Tuart woodland in Yanchep National Park (north of Perth, 31.533°S, 115.683°E) following a high intensity fi re in January 2005, indicate they can be com- patible for bridal creeper. The biological Figure 3. Individual plots shown in a three-dimensional ordination (non- control agent, the bridal creeper rust ex- hausts the fl eshy tuberous root system of metric multi-dimensional scaling). The relative distance between plots bridal creeper (Morin et al. 2002, Turner et (represented by triangles) represent the degree of dissimilarity of native al. 2004). At Yanchep National Park, it ap- plant assemblages between burnt plots at the end of the study period in peared that the fi re, following fi ve years of May 2006. Analysis of similarity revealed no signifi cant difference between biological control, was able to burn most bridal creeper plots (herbicide untreated) or removal plots (herbicide of the senesced root system. In addition treated): Sample statistic (Global R): 0.017, P = 0.274. in November 2005, the biological control

Table 2. Differences in mean (± SE) soil variables between treatments in May 2006A. In 2006, the abundance of bridal creeper differed between plots and therefore the abundance is also documented below. Soil variable Burnt bridal creeper Burnt removal Unburnt bridal FP plots (herbicide treated) plots creeper area Nitrate (mg Kg−1) 5.60 ± 1.57 5.20 ± 0.20 2.60 ± 0.60 3.70B 0.056 Ammonium (mg Kg−1) 5.80 ± 1.46b 2.40 ± 0.24a 2.60 ± 0.24a 5.75B 0.018 Phosphorus (mg Kg−1) 6.00 ± 1.22 3.80 ± 0.20 3.60 ± 0.40 3.02B 0.087 Potassium (mg Kg−1) 165.6 ± 28.9b 83.8 ± 7.0a 101.2 ± 5.8ab 6.06 0.015 Organic carbon (%) 4.29 ± 0.43b 1.87 ± 0.18a 1.64 ± 0.32a 20.02 <0.001 pH 6.86 ± 0.24b 7.62 ± 0.21ab 8.10 ± 0.18a 8.75 0.005 Bridal creeper (shoots m−2) 38.2 ± 7.2C 4.8 ± 1.0C 94.0 ± 9.4 The burnt bridal creeper plots and the unburnt bridal creeper area were not treated with herbicide and the burnt removal plots were treated with herbicide. Different letters after SE relate to signifi cant difference at P = 0.05 based on Tukey multiple comparisons tests. A The analysis of variance models with treatment (n = 5) as the only factor. B Log10 transformation applied. C This abundance relates to the centre 1 × 1 m of the plot, where the soil cores were taken. Plant Protection Quarterly Vol.24(1) 2009 25 agents were observed to have re-colonized CSIRO. Thanks to Frank Anderson, Robin Fox, J. (1984). A comparison of two climb- bridal creeper plants that had survived or Coles, Paul Thomas, Brett Broomell, Vicki ing plant species (one native and one germinated since the fi re. However, more Hawker and Dennis Gannaway for their exotic) at Woodman Point, Western research into this area is needed, given assistance in the fi eld. Comments from Australia. Western Australian Naturalist that fi re is a natural occurrence in most two anonymous reviewers also improved 16, 11-15. Australian ecosystems and weeds that this paper. Fox, M.D. and Fox, B.J. (1986). The suscep- have invaded these native ecosystems are tibility of natural communities to inva- increasingly becoming targets for biologi- References sion. In ‘Ecology of biological invasions: cal control (McFadyen 1998). Adams, R. and Simmons, D. (1991). The an Australian perspective’, eds R.H. Because fi re is a disturbance, it may in- invasive potential of Genista monspessu- Groves and J.J. Burdon, pp. 57-66. (Aus- crease the chances of weed invasion (Fox lana (Montpellier broom) in dry sclero- tralian Academy of Science, Canberra). and Fox 1986, Adams and Simmons 1991, phyll forest in Victoria. Victorian Natu- GenStat. (2003). ‘Release 7.2, Lawes Agri- Milberg and Lamont 1995). However, no ralist 108, 84-89. cultural Trust (Rothamsted Experimen- exotic species benefi ted following the fi re Bell, D.T., Plummer, J.A. and Taylor, S.K. tal Station).’ (VSN International Ltd, or following the control of bridal creeper (1993). Seed germination ecology in Hemel Hempstead). during this study at Meningie. Exotics southwestern Western Australia. The Handreck, K.A. (1997). Phosphorus re- other than bridal creeper only averaged a Botanical Review 59, 24-73. quirements of Australian native plants. maximum of 3.45% cover (Table 1) and na- Briese, D.T. (1996). Biological control of Australian Journal of Soil Research 35, tive species dominated suggesting that fi re weeds and fi re management in pro- 241-89. did not decrease the site’s resilience. Weed tected natural areas: are they compat- Hester, A.J. and Hobbs, R.J. (1992). Infl u- species present at the site but remaining ible strategies? Biological Conservation ence of fi re and soil nutrients on native at low density included Ehrharta calycina 77, 135-41. and non-native annuals at remnant Sm., Brassica tournefortii Gouan and Freesia Byers, J.E., Reichard, S., Randall, J.M., vegetation edges in the Western Aus- sp. Thomson and Leishman (2005) found Parker, I.M., Smith, C.S., Lonsdale, tralian wheatbelt. Journal of Vegetation that fi re did not promote invasion into ar- W.M., Atkinson, I.A.E., Seastedt, T.R., Science 3, 101-8. eas that were non-nutrient enriched. This Williamson, M., Chornesky, E. and Hobbs, R.J. (1991). Disturbance a precursor is supported by Hester and Hobbs’ (1992) Hayes, D. (2002). Directing research to weed invasion in native vegetation. suggestion that fi re does not necessarily to reduce the impacts of nonindig- Plant Protection Quarterly 6, 99-104. increase weed invasion and was a useful enous species. Conservation Biology 16, Hobbs, R.J. and Huenneke, L.F. (1992). management tool in remnant vegetation 630-40. Disturbance, diversity and invasion: in Western Australia. However, the Men- Carr, B. (1996). Bridal creeper at Woodman implications for conservation. Conser- ingie study was undertaken at only one Point – its current status and recom- vation Biology 6, 324-37. site, and therefore further research is rec- mended control strategies. Plant Protec- Kleinjan, C.A. and Edwards, P.B. (1999). ommended, given that the outcomes may tion Quarterly 11, 67-9. A reappraisal of the identifi cation and vary across sites depending on timing of Clarke, K.R. (1993). Non-parametric mul- distribution of Asparagus asparagoides in fi re, fi re intensity, pre-fi re invasion status tivariate analyses of changes in com- southern Africa. South African Journal of and soil quality. munity structure. Australian Journal of Botany 65, 23-31. Fire can be an effective tool for weed Ecology 18, 117-43. Lake, J.C. and Leishman, M.R. (2004). In- management, especially when used in Clarke, K.R. and Warwick, R.M. (2001). vasion success of exotic plants in natu- conjunction with other control techniques ‘Change in marine communities: an ap- ral ecosystems: the role of disturbance, (Willis et al. 2003). However, in highly dis- proach to statistical analysis and inter- plant attributes and freedom from turbed roadside remnants and those areas pretation, second edition.’ (PRIMER-E, herbivores. Biological Conservation 117, that have elevated soil nutrient levels, the Plymouth). 215-26. use of fi re may promote exotic species (for Downey, P.O. (2006). The Weed Impact to Luken, J.O. (1997). Management of plant example see Milberg and Lamont 1995, Native Species (WINS) assessment tool invasions: implicating ecological suc- Thomson and Leishman 2005). The degree – results from a trial for bridal creeper cession. In ‘Assessment and manage- of weed invasion could change depending (Asparagus asparagoides (L.) Druce) and ment of plant invasions’, eds J.O. Luken on fi re frequency and intensity as well as ground asparagus (Asparagus aethiopi- and J.W. Thieret, pp. 133-44. (Springer- season of the burn, for example a spring cus L.) in southern New South Wales. Verlag, New York). versus autumn burn (Hobbs and Huen- Plant Protection Quarterly 21, 109-16. McCaw, W.L. (1998). Regeneration of Aca- neke 1992). This study indicates that fi re Duggin, J.A. and Gentle, C.B. (1998). Ex- cia and Kennedia from soil stored seed can be an important restoration tool to perimental evidence on the importance following an autumn fi re in jarrah (Eu- stimulate the regeneration of native Aus- of disturbance intensity for invasion of calyptus marginata) forest. Journal of the tralian plants and speed up the recovery of L. in dry rainforest-open Royal Society of Western Australia 71, bridal creeper invaded ecosystems, provid- forest ecotones in north-eastern NSW, 1-6. ed that bridal creeper and other weeds are Australia. Forest Ecology and Manage- McFadyen, R.E.C. (1998). Biological con- kept at a low post-fi re density, naturally or ment 109, 279-92. trol of weeds. Annual Review of Entomol- through targeted control. Further research Fisher, J. (1999). The role of the soil seed ogy 43, 369-93. is now needed to confi rm the suitability of bank in bushland management. In Milberg, P. and Lamont, B.B. (1995). Fire fi re across a range of bridal creeper sites ‘Managing our bushland: Proceedings enhances weed invasion on roadside throughout southern Australia. of a Conference about the Protection vegetation in southwestern Australia. and Management of Urban Bushland’, Biological Conservation 73, 45-9. Acknowledgements eds K. Tullis and K. McLean, pp. 134-8 Morin, L., Batchelor, K.L. and Scott, J.K. Coorong District Council kindly pro- (Urban Bushland Council WA Inc, West (2006a). The biology of Australian vided permission to use the reserve. This Perth). weeds. 44 Asparagus asparagoides (L.) project was funded by the Government of Fogarty, G. and Facelli, J.M. (1999). Growth Druce. Plant Protection Quarterly 21, 46- South Australia. Peter Turner was fund- and competition of Cytisus scoparius, an 62. ed through an Australian Postgraduate invasive shrub, and Australian native Morin, L., Neave, M., Batchelor, K. and Award at UWA, a top-up scholarship from shrubs. Plant Ecology 144, 27-35. Reid, A. (2006b). Biological control: a 26 Plant Protection Quarterly Vol.24(1) 2009 promising tool for managing bridal phosphorus. Proceedings of the 9th In- creeper, Asparagus asparagoides (L.) ternational Conference on the Ecology Druce, in Australia. Plant Protection and Management of Alien Plant Inva- Quarterly 21, 69-77. sions – Perth 2007. Morin, L., Willis, A.J., Armstrong, J. and Turner, P.J., Scott, J.K. and Spafford Jacob, Kriticos, D. (2002). Spread, epidemic H. (2006). Barrier to restoration: the development and impacts of the bridal decomposition of bridal creeper’s root creeper rust in Australia: summary of system. Proceedings of the 15th Aus- results. Proceedings of the 13th Aus- tralian Weeds Conference’, eds C. Pres- tralian Weeds Conference, eds H. Spaf- ton, J.H. Watts and N.D. Crossman, pp. ford Jacob, J. Dodd and J.H. Moore, pp. 827-30. (Weed Management Society of 385-8. (Plant Protection Society of WA South Australia, Adelaide). Inc, Perth). Turner, P.J., Spafford, H. and Scott, J.K. (in Raymond, K. (1995). The autecology of press). The ecological barriers to the bridal creeper: how does it work? In recovery of bridal creeper (Asparagus ‘Weeds of conservation concern’, eds asparagoides (L.) Druce) infested sites: D. Cooke and J. Choate, pp. 17-21. (De- impacts on vegetation and the potential partment of Environment and Natural increase in other exotic species. Austral Resources, Animal and Plant Control Ecology. Commission, Adelaide). Turner, P.J. and Virtue, J.G. (2006). An Raymond, K. (1996). The ecology of bridal eight-year removal experiment meas- creeper in south-eastern Australia. Plant uring the impact of bridal creeper (As- Protection Quarterly 11, 47. paragus asparagoides (L.) Druce) and the Shea, S.R., McCormick, J. and Portlock, potential benefi t from its control. Plant C.C. (1979). The effect of fi res on regen- Protection Quarterly 21, 79-84. eration of leguminous species in the Willis, A.J., McKay, R., Vranjic, J.A., Kilby, northern jarrah (Eucalyptus marginata M.J. and Groves, R.H. (2003). Compar- Sm) forest of Western Australia. Aus- ative seed ecology of the endangered tralian Journal of Ecology 4, 195-205. shrub, and a threaten- Sorensen, B. and Jusaitis, M. (1995). The ing weed, bridal creeper: smoke, heat impact of bridal creeper on an endan- and other fi re-related germination cues. gered orchid. In ‘Weeds of conservation Ecological Management and Restoration 4, concern’, eds D. Cooke and J. Choate, 55-65. pp. 27-31. (Department of Environment Willis, A.J., Morin, L., Moore, P.H.R. and and Natural Resources, Animal and Groves, R.H. (2004). Potential for popu- Plant Control Commission, Adelaide). lation recovery of an endangered na- Thomson, V.P. and Leishman, M.R. (2005). tive plant by controlling bridal creeper Post-fi re vegetation dynamics in nu- with rust. Proceedings of the XI Interna- trient-enriched and non-enriched scle- tional Symposium of Biological Control rophyll woodland. Austral Ecology 30, of Weeds, 2003, eds J.M. Cullen, D.T. 250-60. Briese, D.J. Kriticos, W.M. Lonsdale, L. Thorp, J.R. and Lynch, R. (2000). ‘The de- Morin and J.K. Scott, pp. 483. (CSIRO termination of weeds of national sig- Entomology, Canberra). nificance.’ (National Weeds Strategy Executive Committee, Launceston, Tas- mania). Turner, P.J., Morin, L., Williams, D. and Kriticos, D. (2004). Interactions between two weed biological control agents, an insect and pathogen, and the response of their host. Proceedings of the 14th Australian Weeds Conference, eds B.M. Sindel and S.B. Johnson, pp. 398. (Weed Society of New South Wales, Sydney). Turner, P.J., Scott, J.K. and Spafford, H. (2008). Implications of successful bio- logical control of bridal creeper (Aspar- agus asparagoides (L.) Druce) in south west Australia. Proceedings of the 16th Australian Weeds Conference, eds R.D. van Klinken, V.A. Osten, F.D. Panetta and J.C. Scanlan, pp. 390-2. (Queens- land Weeds Society, Brisbane). Turner, P.J., Scott, J.K. and Spafford, H. (submitted). The ecological barriers to the recovery of bridal creeper (As- paragus asparagoides (L.) Druce) infest- ed sites: elevated levels of available