Over-abundance of native herbivores is a problem in many forests worldwide. The abundance of native macropod wallabies is extremely high at Booderee National Park (BNP) in south-eastern Australia. This has occurred because of the reduction of exotic predators through an intensive baiting program, coupled with the absence of other predators. The high density of wallabies at BNP may be inhibiting the recruitment of many plant species following fire-induced recruitment events. We experimentally examined the post-fire response of a range of plant species to browsing by wallabies in a forest heavily infested with the invasive species, bitou bush Chrysanthemoides monilifera. We recorded the abundance and size of a range of plant species in 18 unfenced (browsed) and 16 fenced (unbrowsed) plots. We found the abundance and size of bitou bush was suppressed in browsed plots compared to unbrowsed plots. Regenerating seedlings of the canopy or middle storey tree species Eucalyptus pilularis, Acacia implexa, Allocasuarina littoralis, Breynia oblongifolia and Banksia integrifolia were either smaller or fewer in number in grazed plots than treatment plots as were the vines Kennedia rubicunda, Glycine tabacina and Glycine clandestina. In contrast, the understorey fern, Pteridium esculentum increased in abundance in the browsed plots relative to unbrowsed plots probably because of reduced competition with more palatable angiosperms. Twelve months after plots were installed the community structure of the browsed and unbrowsed plots was significantly different (P = 0.023, Global R = 0.091). The relative abundance of C. monilifera and P. esculentum contributed most to the differences. We discuss the possible development of a low diversity bracken fern parkland in Booderee National Park through a trophic cascade, similar to that caused by overabundant deer in the northern hemisphere. We also discuss its implications for broad scale fox control in southern Australian forests.
Citation: Dexter N, Hudson M, James S, MacGregor C, Lindenmayer DB (2013) Unintended Consequences of Invasive Predator Control in an Australian Forest: Overabundant Wallabies and Vegetation Change. PLoS ONE 8(8): e69087. https://doi.org/10.1371/journal.pone.0069087
Editor: Michael Somers, University of Pretoria, South Africa
Received: September 17, 2012; Accepted: June 11, 2013; Published: August 21, 2013
Copyright: © 2013 Dexter et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: Funding was provided by Parks Australia, and the Australian Research Council. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The problem of overabundant herbivores changing ecosystem structure through grazing or browsing is a well recognised worldwide phenomenon in terrestrial and marine ecosystems       . For example, this is occurring in the hardwood forest ecosystems of much of Europe and North America, where the removal of predators and restrictions on hunting by humans has resulted in a substantial increase in deer populations leading to major impacts on forest structure and biodiversity , . These heavily browsed temperate forest communities become fern parkland comprised of mature and senescent trees with a floristically depauperate understorey dominated by unpalatable ferns , , . The impacts of persistent heavy browsing can extend through trophic cascades to influence not only plants but also invertebrates , birds , and soil micro-organisms .
In the Eucalyptus forests of southern Australia, relatively little attention has been paid to the ecological role of browsing mammals in shaping forest structure and floristics. Instead, wildfire has been considered to be the main form of disturbance influencing the composition and structure of these forests , , . However, studies by Leigh and Holgate  and Parsons et al.  examined the response of a range of plant species to native mammal herbivory, (principally macropods: kangaroos and wallabies) in forest and woodland habitats following fire in southern Australia. Both these studies stressed that following fire, herbivores had a substantial effect on the recruitment of a wide range of native plants. Parallel to these investigations are studies in southern Australia on the impact of browsing macropods on the regeneration of commercially valuable Eucalyptus species after forestry operations ,  and revegetation attempts during mine site rehabilitation , . This research has revealed that mammalian browsing has major impacts on plant regeneration and plant species composition. These studies, together with literature on the impact of browsing ungulates in northern hemisphere forests , , suggest that herbivory may be an under-recognised factor influencing the composition and structure in Australian forests.
Despite their probable importance in shaping the structure of Australian forests, many small browsing macropods have suffered significant population decline, or even extinction due to predation by the introduced red fox (Vulpes vulpes) ,  . Indeed, a small forest macropod, the Tasmanian pademelon (Thylogale billardierii), has probably been extirpated from mainland Australia by fox predation. Yet it is a major pest in wood production forests in Tasmania where foxes are extremely rare . Many fox control programs have been initiated in southern Australia in an attempt to reverse macropod decline , , , . In most instances, these control programs have resulted in a large-scale increase in the abundance of terrestrial mammals, including wallabies , , , , . This increase is evident even in large macropods such as eastern grey kangaroos Macropus giganteus .
Since 1999, Booderee National Park (BNP) on the south-east coast of Australia, has maintained an intensive fox control program that has resulted in the increase in abundance of a number of native mammal species , . There has been an increase in the abundance of macropods of approximately 10 fold between 2003 and 2009. The swamp wallaby (Wallabia bicolor) is a specialist browsing macropod  and is the most common macropod species at BNP. It is also an important component of the diet of foxes in BNP , . Less common mammalian herbivores present at BNP are the eastern grey kangaroo (Macropus giganteus), red necked wallaby (Macropus rufogriseus), common brushtail possum (Trichosurus vulpecula), common ringtail possum (Pseudocheirus peregrinus) and bush rat (Rattus fuscipes) . European rabbits (Oryctolagus cuniculus) are also present but their distribution is very patchy and their numbers very low and were never observed on the study site. Coincident with the decrease in fox abundance and the increase in wallaby abundance, observations of the flora at BNP suggest that few seedlings have established after fire and that revegetation programs following fire have failed with most individual plants experiencing mortality from browsing.
In this paper, we developed a new conceptual model (Figure 1) of possible long-term changes in the vegetation at BNP whereby the intensive control of foxes has led to a substantial increase in the abundance of wallabies. In turn, this will negatively affect the recruitment of palatable plant species following fire. We also predicted that the recruitment of unpalatable plants will be enhanced through reduced competition with more palatable species. Therefore, we designed an exclusion experiment to determine whether browsing by macropods was reducing the abundance and size of palatable plants and increasing the abundance and size of unpalatable plants. The browsing exclusion experiment was conducted over one year, following a burn to control the environmental weed; bitou bush, Chrysanthemoides monilifera rotundifolia in a mixed blackbutt Eucalyptus pilularis, dry sclerophyll forest. Forests dominated by E. pilularis are common throughout coastal eastern Australia .
Importantly, our study differs from earlier investigations of the impact of browsing herbivores where workers assumed that their study systems were in equilibrium. By contrast, in this study, we know that herbivore abundance is increasing and that the system was possibly shifting to a new equilibrium. Therefore, we wished to be able to make robust predictions about the possible future state of forested ecosystems at BNP and more broadly for other forested areas where fox control is being undertaken.
Materials and Methods
We conducted this study at Booderee National Park, a national park in south-eastern Australia approximately 6500 ha in area (approximate midpoint is 35°10′S latitude, 150°40′E longitude). The area has a temperate maritime climate with an average rainfall of 1225 mm per annum spread relatively evenly over the year. BNP is unusual in having a high diversity of habitats ranging from rainforest to sedgeland in a relatively small area , .
Vegetation communities at BNP have been mapped and classified in detail by Taws . Approximately one third of BNP has been invaded by C. monilifera. Chrysanthemoides monilifera is a perennial evergreen shrub in the family Asteraceae (a native of South Africa), and major environmental weed of dune and coastal forest in southern Australia . At BNP a range of habitats is affected by C. monilifera invasion, but this invasive plant is most common in coastal dune shrubland formerly dominated by Acacia sophorae and Leptospermum laevigatum where C. monilifera was planted in the 1960s to stabilise the unstable dune system . From the dune shrubland, C. monilifera invaded E. botryoides and E. pilularis dry sclerophyll forest adjacent to the dune system. In these communities, C. monilifera has formed a dense 2 to 3 m high monoculture. Aerial spraying of C. monilifera bush commenced in the late 1990s with the study site sprayed in June 2005 and partially resprayed in May–June 2006 prior to burning.
Our conceptual model of the hypothesised species responses is shown in Figure 1. We predicted that following the control burn, there would be proportionally more palatable plants in unbrowsed plots and that there would be proportionally more unpalatable plants in the browsed plots. However, predicting the response to browsing of plants that have different life forms using simple traits is difficult . Nonetheless, based on life history characteristics and data in the literature on the susceptibility of different plant genera to grazing or browsing, we made predictions about the impact of herbivory on the abundance and size of 10 of the more common species.
For the four most common tree species, we predicted that seedlings or ramets of blackbutt Eucalyptus pilularis, coastal banksia Banksia integrifolia, lightwood Acacia implexa and black she-oak Allocasuarina littoralis were likely to be more abundant and taller in unbrowsed than browsed plots given the known susceptibility of these genera to browsing , , , , . We also included coffee bush Breynia oblongifolia in the analysis as our initial observations indicated a heavy level of grazing pressure on this mid-storey tree. Scurr et al  demonstrated that the distribution of boneseed Chrysanthemoides monilifera ssp. monilifera in Tasmania was constrained by domestic and wild browsers. In light of this, we predicted the conspecific bitou C. monilifera ssp rotundata would also be less abundant and smaller in size within browsed plots. Some plants in the family Fabaceae have relatively high concentrations of nitrogen making them attractive to herbivores . Therefore, we predicted that the twining legumes Kennedia rubicunda and Glycine spp would be heavily browsed and more abundant and larger in treatment plots than the controls .
In contrast, for two species (eastern nightshade Solanum pungetium and bracken fern Pteridium esculentum) we predicted that there would be an increase in size and abundance in browsed plots. For the highly prickly S. pungetium, we predicted an increase in size or abundance, as prickliness in Australian Solanum may be an adaptation to deter wallaby browsing . As a dense understorey of unpalatable ferns is a common outcome of intensive browsing in northern hemisphere forests, we predicted an increase in the abundance and/or size of P. esculentum. Pteridium esculentum is relatively toxic and unpalatable to mammalian herbivores including macropods  due to a cyanogenic glucoside , .
Finally, we included two very common species in our analysis which respond rapidly to fire or disturbance; blady grass Imperata cylindrica and black nightshade Solanum nigrum. Imperata cylindrica is a tough rhizatomous grass that tends to become dominant with frequent burning  and S. nigrum a cosmopolitan ruderal, annual, weed . We made no specific predictions about these two species as data on their palatability and toxicity are equivocal.
In August 2007, we established 18 unbrowsed plots (excluding macropod browsing) and 18 browsed plots (allowing macropod browsing). Unbrowsed plots consisted of a 3×3 m enclosure of 2.5 cm mesh wire 150 cm high, with the wire supported by strained metal star pickets at each corner. To access a fenced plot, we installed a ∼60 cm wide door consisting of a secured mesh flap. The wire mesh was buried to a depth of 20 cm to prevent access by burrowing mammals. Browsed plots were 3×3 m with star pickets on each corner but no mesh. Plots were not chosen at random but selected to maximise the probability that they would be burnt (i.e. high mass of dead C. monilifera canes). An effort was made to pair browsed and unbrowsed plots by separating them by at least 20 m. However, because of the heterogeneous dispersion of fuel sufficient to sustain a fire, not all plots could be paired. Therefore, the pairs of plots were not treated as blocks in the sense of a randomized block design experiment. It was assumed that both species of possums and bush rats would be able to access the unbrowsed plots.
The study sites were located in an area of forest that was part of a spray-burn-spray treatment for C. monilifera. This treatment involves spraying C. monilifera with the herbicide (glyphosate), allowing the plants to die and cure, and then burning the cured canes. This stimulates a mass germination by C. monilifera which can significantly reduce the seedbank. Before these seedlings can flower and set seed, the infestation is re-sprayed , . An earlier study has indicated that recruitment of native plants was better following a spray-burn-spray treatment than either spraying alone, or spraying and burning .
Test burns indicated that when conditions were suitable for ignition of the cured canes, combustion of clumps was briefly intense for 5–10 minutes. Burning was conducted over two days in October 2007. We burnt a total of 226 ha. Combustion was rapid and relatively intense with flame heights up to 5 metres in tall (2–3 m) stands of cured C. monilifera. All 18 browsed sites were completely burnt, but two of the 18 unbrowsed sites were not burnt and were not subsequently used in the analysis reported here.
We began bi-annual fox control in 1999 using lethal 1080 baits placed at 1-km intervals along a network of trails throughout BNP (for a full description of the methods see Dexter et al . Foxes were controlled as part of a long term program to protect biodiversity in the Park. The control of foxes using lethal 1080 baits is a required action under the Booderee National Park Plan of Management (1999): “5.10.10 Maintain a control program for foxes and monitor the program's effectiveness, amending control methods as required .” The Booderee National Park Plan of Management is the legal frame work that the Australian Federal Government (through Parks Australia) uses to manage the Park. As such an ethics approval is not required to control foxes.
Baiting was increased to monthly in 2003. Baits were checked after 14 days and the number of baits taken by foxes expressed as a percentage of the total number of baits available. Bait take is a rough index of abundance suitable only for measuring large scale changes in abundance over large areas but it has been used successfully to monitor changes in fox relative abundance in other studies  
We counted macropods by repeated spotlighting along 109 permanent 100 m long transects established throughout BNP (for a full description of the methods see Lindenmayer et al ). Counts were not undertaken on nights of poor weather. Experienced observers completed the spotlighting surveys annually from 2003 to 2007, and biennially from 2007. Macropods were usually detected by sound of their footfall when they were disturbed. Where possible, the species of macropod was identified as swamp wallaby, eastern grey kangaroo or red necked wallaby. Wallabies were monitored under the Australian National University Animal Ethics Protocol number C.RE.60.09.″
One month after the fire, all the plots were surveyed. The following attributes were recorded; the number of individuals, percentage cover (in five classes <5%; 5–25%, 25–50%; 50–75%; 75–100%), height range (tallest and shortest individual for all species), and diameter range (widest and narrowest individual for species wider than tall) for all seedling and resprouting species. Mature trees were not included in this study. These surveys were repeated at two monthly intervals until November 2008 giving a total of seven repeated surveys.
Plant abundance data were the counts of the 12 individual species in each plot. The hypothesis that browsing affected species abundance was tested using the mixed models option with the diagonal offset covariance structure for the repeated-measure in the statistical program SPSS. Treatment (browsed, unbrowsed) was entered as a fixed factor, with sample session (1–7) entered as a repeated-measure and sampling session by treatment entered as an interaction term. Five species were not detected until the second sampling session (A. littoralis, B.integrifolia, K. rubicunda, S. nigrum, and, S. pungetium). For these species, sessions 2–7 were entered as the repeated-measure. For species with a highly right-skewed abundance distribution, data were transformed by the fourth root. Mixed models are better at handling the more complex covariance structure values than traditional repeated measure ANOVAs, including missing values . They are also relatively robust to non-normality .
For the analysis of plant size, the median height in centimetres for nine of the 12 species, recorded in both unbrowsed and browsed plots were also entered into a mixed model. For Glycine spp, K. rubicunda and S. pungetium the median width in cm was used as this dimension tends to be greater than height for these species.
To determine whether macropod browsing influenced the overall community structure of the regenerating vegetation, we compared the differences between browsed and unbrowsed plots using an analysis of similarity (ANOSIM). We used the November 2008 data for plant abundance for all species detected as these represented the final and most diverse community sample. The ANOSIM was used after square root transforming the data to emphasise the importance of rarer species. To determine which species contributed most strongly to differences between browsed and unbrowsed plots, we entered all the transformed data into a Similarity Percentage Analysis (SIMPER). Both these analyses were completed using PRIMER v6 .
We extended the analysis of community structure to include an index of species mass. We did this by repeating the ANOSIM but using a composite value consisting of the number of each species in each plot in November 2008 multiplied by the median size (height or width). As with the ANOSIM for numbers of individuals, the data were square root transformed to emphasise the importance of rarer species. These data also were analysed with a SIMPER.
Fox abundance, as indexed by bait take, declined precipitously from 24% of baits taken in 1999 before starting to recover to 32% in 2003 when monthly baiting commenced . This brought bait take to a consistently low level (less than 1% of baits taken)  However, it is unlikely that true abundance of foxes declined as much as the bait take, as foxes continue to be observed in low numbers.
Results from spotlight transect counts revealed a 10 fold increase in the abundance of wallabies. A total of 5 individuals were detected in 2003 compared to 60 in 2009 . The majority of observations were of swamp wallabies (77.1% of observations), with fewer eastern grey kangaroos (19.4% of observations) and red-necked wallabies (3.5% of observations). Over the same period common brushtail possums increased while common ringtail possums declined 
Browsing significantly reduced the mean number of seedlings and resprouts for 6 of the 12 species surveyed (Table 1, Figure 2). The results accorded largely with our predictions about which species would be significantly affected by browsing; Chrysanthemoides (P = 0.038), Eucalyptus (P = 0.006), Banksia (P<.0001), Allocasuarina (P = 0.001), Breynia (P = 0.036) and Glycine (P = 0.001) were all significantly reduced in abundance in the browsed plots compared to the unbrowsed plots. However, the differences in abundance for Eucalyptus and Banksia had almost disappeared by the last count. The other main canopy species, Acacia implexa, was not significantly affected by macropod browsing. There were no significant interaction effects detected between treatment and sample session for any species examined.
On average, the abundance of the prickly nightshade Solanum pungetium was higher in browsed plots (0.875 to 0.65) but was not significantly different between grazed and unbrowsed plots (P = 0.52). However, there were significantly more P. esculentum ramets in browsed plots than unbrowsed plots (P<0.001). The abundance of the rapidly recolonising species I. cylindrica and S. nigrum were not affected by browsing.
The overall pattern was for larger plants to occur in unbrowsed plots compared to browsed plots. C. monlifera (P<0.0001), E. pilularis (P<0.0001), A.implexa (P<0.0001), and B. Integrifolia (P<0.0001) S. nigrum, (P<0.0001) and I. cyclindrica (P<0.001) were significantly taller in unbrowsed plots while K. rubicunda (P<0.0001) and Glycine (P = 0.003) were wider in unbrowsed plots (Table 2, Figure 3).
There were strong interactive effects between sample session and treatment for C. monlifera (P<0.0001), E. pilularis (P<0.0001), A.implexa (P<0.0001), and B. Integrifolia (P<0.0001) S. nigrum, (P<0.0001) and I. cyclindrica (P<0.001). The significant interactive effects indicated that in the unbrowsed plots plants continued to grow during the study but that in the browsed plots most plants remained stunted by browsing.
There was no significant treatment effect for S. pungetium but the interaction term between treatment and sample session was significant (P = 0.045). This interactive effect is shown in Figure 3 and indicates S. pungetium was much larger in browsed than unbrowsed plots for most of the study. However, by the end of the study, plants in unbrowsed plots had died and only small seedlings remained in browsed and unbrowsed plots.
A total of 70 species was detected during the vegetation surveys with 62 species recorded in the unbrowsed plots and 62 species detected in the browsed plots. Species detected only in browsed plots were Centaurium erythraea, Ehrharta erecta, Eustrephus latifolius, Lagenifera stipitata, Persoonia linearis, Pratia purpurascens and Wahlenbergia gracilis. These species tended to be small herbs or grasses (except P. linearis and E.latifolius). Similarly, the small (<10 cm) succulent forb Crassula sieberana was detected in only one unbrowsed plot and in six browsed plots. The species recorded only in unbrowsed plots were Acacia ulicifolia, Billardiera scandens, Gahnia clarkei, Glochidion ferdinandi, Omolanthus populifolius, Pelargonium australe, Poa labillardieri,and Trema tomentosa. Three of these species (G. ferdinandi, O. populifolius and T. tomentosa) were the seedlings of large shrubs or small trees that commonly form a mid-storey in coastal sclerophyll forest and rainforest.
The results of ANOSIM revealed significant differences in species composition between browsed and unbrowsed plots (Global R = 0.091, P = 0.023). The species contributing up to 50% of the dissimilarity between browsed and unbrowsed plots are listed in Table 3. The ANOSIM of species mass showed a larger difference between browsed and unbrowsed plots (Global R = 0.195. P<0.001) with species contributing up to 50% of dissimilarity between browsed and unbrowsed plots listed in Table 4. The contribution of the two most important species C. monilifera and P. esculentum to the observed differences in community structure can be seen by examining Figure 4 which shows the distribution of percentage cover classes of both species. It is clear that P. esculentum was more likely to be dominant in the unfenced plots than the fenced plots (Table 4).
Our study demonstrated that following fire at sites protected from browsing, a wide variety of palatable plants were more common or larger in unbrowsed sites than in browsed sites. Our study also demonstrated that two species of unpalatable plants benefited from the reduced competition with palatable plants in browsed sites. Of broader theoretical and practical importance, our study demonstrated the unforseen impacts on vegetation of a program to control an exotic carnivore. This pattern is similar to that seen on sub-Antarctic islands where feral cat Felis catus control led to an irrruption in exotic European rabbit Oryctolagus cuniculus numbers, and in turn, resulted in serious negative impacts on native vegetation .
Many Northern Hemisphere studies have been conducted in the ecological context of high deer densities that have existed for up to 50 years , , , hence, researchers were already examining heavily altered ecosystems. In Australia studies on the impacts of macropod browsing on managed forests and mine site revegetation predation by either foxes or dingoes is likely to have had a constraining effect on the abundance of macropods , , .
The lack of large mammalian predators is a novel circumstance in forests in eastern Australia.Prior to the introduction of foxes, wallabies would have been hunted by both Aboriginal people and dingoes (Canis lupus dingo) . We therefore believe that at BNP, browsing pressure is now much higher than in the recent past, although the actual level of browsing pressure is unknown with certainty. The significance of our study is its timing at the beginning of a phase of very high herbivore abundance.
We have assumed that BNP is unusual because heavy browsing has resulted from intensive fox control. However, intensive fox control is becoming more common. Indeed, 100,000s of ha of forest in south-eastern and south-western Australia now under intensive fox control , . Our experiment shows that an unexpected outcome of such control efforts may be major structural and floristic change to these forests due to increases in wallaby abundance.
We found that overstorey species (E. pilularis, B. integrifolia, A. implexa and A. litoralis) were either reduced in abundance, size, or both due to wallaby browsing. However, this difference may not carry through to maturity given processes such as self thinning, as this is very common in Eucalyptus species following fire . The lack of significant interactive effects between sample session and treatment on abundance indicates that despite many plants being removed by browsing, differences in abundance remained consistent through the course of the study. This is because most species continued to germinate and resprout, thereby adding new recruits to both browsed and unbrowsed sites at roughly equal rates. Profound effects of macropod browsing on both Banksia and Eucalyptus regeneration have previously been documented , , . Compared to understorey species, trees are buffered against death and damage from browsing by eventually growing higher than herbivores can browse . However, some trees may become permanently stunted and remain within the browsing zone if browsing pressure is strong enough . This may be the case for species like B. integrifolia and E. pilularis, which had significant interactive terms in the analysis indicating that while plants in unbrowsed plots continued to grow, plants in browsed plots hardly changed height over the entire course of the study (Figure 2). Despite the relative immunity of canopy trees to browsing, evidence from North America indicates that the composition of canopy hardwoods can be changed from more palatable species to less palatable species as a result of intense browsing pressure .
Palatable understorey species such as B. oblongifolia, K. rubicunda and Glycine spp. remain exposed to browsing throughout their life cycle and may be susceptible to extirpation . Kennedia spp. have persistent seed banks that are stimulated to germinate by fire , so they can potentially survive long periods even if all adult plants are consumed by browsers. In contrast, S. pungetium was substantially larger in browsed plots than unbrowsed plots for most of the study. This may be because this sprawling unpalatable annual forb had reduced competition from other more palatable plants in the browsed plots. There was a tendency for some very small (<5 cm) grasses and forbs to be present or more abundant only in browsed plots. This is most likely due to a lack of larger more vigorous competitors. Shorter growing species tend to increase compared to taller growing species, under moderate grazing pressure, in structurally simple grasslands, although this trend may not hold at higher grazing pressure , .
The two species that contributed most to differences between browsed and unbrowsed plots were C.monilifera and P. esculentum. This occurred in both our analyses of abundance and mass. Chrysanthemoides monilifera dominated the unbrowsed plots while P. esculentum dominated the browsed plots. This is despite P. esculentum being a dietary item for swamp wallabies . However, di Stefano and Newell  showed that bracken was not a preferred food of swamp wallabies. Subtle differences in phenolics within plant families  and within individual genera  can have profound impacts on the browsing pressure to which individual plant species are exposed. The substantial impact of browsing on C. monilifera is in accord with the finding of Scurr et al  that browsing by macropods and domestic livestock limited the distribution of the conspecific boneseed Chrysanthemoides monilifera ssp. monilifera in Tasmania. In our study, C. monilifera did not grow higher than 1 m in the browsed plots, compared to unbrowsed plots where it grew up to 2 m. Furthermore, in only one browsed plot did C. monilifera have a percentage cover of more than 5%. Hence, its ability to exert its strong competitive suppressive impact on regenerating native flora was greatly reduced by browsing, although it would still be able to suppress competition through the allelopathic properties of its roots .
Pteridium esculentum was more abundant in the browsed plots than unbrowsed plots, possibly because its toxicity not only discouraged browsing but also its allelopathy reduced competition from C. monilifera and the more diverse angiosperm flora in unbrowsed plots. The increased dominance of P.esculentum is consistent with northern hemisphere studies which demonstrate an increase in ferns with high levels of deer browsing , . In Australia, P. esculentum has been shown to have a substantial negative competitive effect on eucalypt seedling survival , so there may be an additive negative effect on eucalypt survival in the browsed plots. In North American studies, ferns were shown to exacerbate the effect of deer browsing by limiting the germination of a range of angiosperms .
Our study points to a larger problem of macropod browsing beyond wood production forests and mine site regeneration in the management of Australian forests. In our study part of a naturally occurring community of plants was altered. This study suffers the limitations of other exclosure studies in that the treatment represents an extreme case of the complete exclusion of terrestrial browsers . A further limitation maybe that increased macropod mortality during a large scale wildfire may mean less intensive post fire browsing. It is also limited to one particular ecosystem, and ecosystem infested with a major environmental weed. This may make inferences about more pristine forests difficult. However, our work does point the way for future studies of vegetation change where large scale exotic predator manipulation may be considered for environmental benefit.
We thank Shane Sturgeon, Tony Carter, Darren Brown, and Gavin McLeod for constructing the unbrowsed plots. We thank Sharon Blazier, Shane Sturgeon and Louise Oliver for assistance in sampling the flora and Stig Pedersen for identification of specimens. This paper was improved with comments from Paul Meek and Louise Oliver.
Conceived and designed the experiments: ND MH SJ. Performed the experiments: ND MH SJ CM DL. Analyzed the data: ND . Wrote the paper: ND.
- 1. Paine RT, Vadas RL (1969) The effect of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnol Oceanogr 14: 710–719.
- 2. Waller DM, Alverson WS (1997) The White-Tailed Deer: A Keystone Herbivore. Wildl Soc Bul 25: 217–226.
- 3. Coulson G (1998) Management of overabundant macropods - are there conservation benefits? In: Austin A, Cowan P editors. Managing Marsupial Abundance for Conservation Benefits. Issues in Marsupial Conservation and Management, Occasional Papers of the Marsupial CRC No. 1, Sydney, 37–48.
- 4. Ripple WJ, Larsen EJ, Renkin RA, Smith DW (2001) Trophic cascades among wolves, elk and aspen on Yellowstone National Park's northern range. Biol Cons 102: 227–234.
- 5. Balfour D, Dublin HT, Fennessy J, Gibson D, Niskanen L, et al. (2007) Review of Options for Managing the Impacts of Locally Overabundant African Elephants. Gland, Switzerland, IUCN, 80.
- 6. Bråthen KA, Ims RA, Yoccoz NG, Fauchald P, Tveraa T, et al. (2007) Induced Shift in Ecosystem Productivity? Extensive Scale Effects of Abundant Large Herbivores. Ecosystems 10: 773–789.
- 7. Vanak AT, Shannon G, Thaker M, Page B, Grant R, et al. (2012) Biocomplexity in large tree mortality: interactions between elephant, fire and landscape in an African savanna. Ecography 35: 315–321.
- 8. Kirby KJ (2001) The impact of deer on the ground flora of British woodland. Forestry 74: 219–230.
- 9. Nomiya H, Suzuki W, Kanazashi T, Shibata M, Tanaka H, et al. (2002) The response of forest floor vegetation and tree regeneration to deer exclusion and disturbance in a riparian deciduous forest, central Japan. Plant Ecol 164: 263–276.
- 10. Horsley SB, Stout S, de Calesta DS (2003) White-tailed deer impact on the vegetation dynamics of a northern hardwood forest. Ecol Appl 13: 98–118.
- 11. Allombert S, Stockton S, Martin J-L (2005) Natural Experiment on the Impact of Overabundant Deer on Forest Invertebrates. Cons Biol 19: 1917–1929.
- 12. De Calesta DS (1994) Effect of white-tailed deer on songbirds within managed forests in Pennsylvania. J Wildl Manage 58: 711–717.
- 13. Wardle DA, Bardgett RD (2004) Human-induced changes in large herbivorous mammal density: the consequences for decomposers. Front Ecol Environ 2: 145–153.
- 14. Attiwill PM (1994) Ecological disturbance and the conservative management of eucalypt forests in Australia. Forest Ecol Manage 63: 301–346.
- 15. Gill AM, Catling PC (2002) Fire regimes and biodiversity of forested landscapes of southern Australia. In: Bradstock RA, Williams, JE and Gill AM editors Flammable Australia: Fire Regimes and Biodiversity of a Continent. Cambridge University Press, 351–369.
- 16. Bradstock R, Gill M, Williams RJ (2012) Flammable Australia Fire Regimes, Biodiversity and Ecosystems in a Changing World. CSIRO PUBLISHING
- 17. Leigh JH, Holgate MD (1979) The response of the understorey of forests and woodlands of the Southern Tablelands to grazing and burning. Aust J Ecol 4: 25–45.
- 18. Parsons MH, Rafferty CM, Lamont BB, Dods K, Fairbanks M (2007) Relative effects of mammal herbivory and plant spacing on seedling recruitment following fire and mining. BMC Ecology 7: 13.
- 19. Wilkinson GR, Neilsen WA (1995) Implications of early browsing damage on the long term productivity of eucalypt forests. Forest Ecol Manage 74: 117–124.
- 20. Di Stefano J (2005) Mammalian browsing damage in the Mt. Cole State forest, southeastern Australia: analysis of browsing patterns, spatial relationships and browse selection. New Forests 29: 43–61.
- 21. Parsons MH, Koch J, Lamont BB, Vlahos S, Fairbanks M (2006) Planting density effects and selective herbivory by kangaroos on species used in restoring forest communities. Forest Ecol Manage 229: 39–49.
- 22. Parsons MH, Lamont BB, Koch J, Dods K (2007) Disentangling competition, herbivory, and seasonal effects on young plants in newly restored communities. Restor Ecol 15: 250–262.
- 23. Frelich LE, Lorimer CG (1985) Current and predicted effects of long-term deer browsing in hemlock forests in Michigan, U.S.A. Biol Conserv 34: 99–120.
- 24. Kinnear JE, Onus ML, Sumner NR (1998) Fox control and rock wallaby population dynamics – II. An update. Wildl Res 25: 81–88.
- 25. Kinnear JE, Sumner NR, Onus ML (2002) The red fox in Australia: an exotic predator turned biocontrol agent. Biol Cons 108: 335–359.
- 26. Saunders GR, Gentle MN, Dickman CR (2010) The impacts and management of foxes Vulpes vulpes in Australia. Mammal Review 40: 181–211.
- 27. Orell P (2003) Western Shield, fauna monitoring and staff training, 1996–2003. A paper prepared for the Western Shield Review, Conservation and Land Management, Perth.
- 28. Dexter N, Murray A (2009) The impact of fox control on the relative abundance of forest mammals in East-Gippsland, Victoria. Wildl Res 36: 252–261.
- 29. Banks PB, Newsome AE, Dickman CR (2000) Predation by red foxes limits recruitment in populations of eastern grey kangaroos. Austral Ecol 25: 283–291.
- 30. Dexter N, Meek P, Moore S, Hudson M, Richardson H (2007) Population Responses of Small and Medium Sized Mammals to Fox Control at Jervis Bay, South-eastern Australia. Pacific Conserv Biol 13: 283–292.
- 31. Lindenmayer DB, MacGregor C, Welsh A, Donnelly C, Crane M, et al. (2008) Contrasting mammal responses to vegetation type and fire. Wildl Res 35: 395–408.
- 32. Dexter N, Hudson M, Carter T, MacGregor C (2011) Habitat-dependent population regulation in an irrupting population of long-nosed bandicoots (Perameles nasuta) . Austral Ecol 36: 745–754.
- 33. Di Stefano J, Newell GR (2008) Diet selection by the swamp wallaby (Wallabia bicolor): feeding strategies under conditions of changed food availability. J Mammal 89: 1540–1549.
- 34. Meek PD, Triggs B (1998) The food of foxes, dogs and cats on two peninsulas in Jervis Bay, New South Wales,. P Linn Soc NSW 120: 117–127.
- 35. Roberts MW, Dexter N, Meek PD, et al. (2006) Does baiting influence the relative composition of the diet of foxes? Wildl Res 33: 481–488.
- 36. Boland D, Brooker I, McDonald MW (2006) Forest Trees of Australia, CSIRO Publishing, Melbourne.
- 37. Cho G (1995) The Jervis Bay environment. In: Jervis Bay. A Place of Cultural, Scientific and Educational Value. Cho G, Georges G, Stoutjesdijk R editors, Kowari 5 Australian Nature Conservation Agency, Canberra pp. 3–8.
- 38. Taws N (1998) Fire and vegetation management in Jervis Bay Territory. Report, Environment Australia, Canberra.
- 39. Winkler MA, Cherry H, Downey PO (2008) Bitou Bush Management Manual: current management and control options for bitou bush (Chrysanthemoides monilifera ssp. rotundata) in Australia. Department of Environment and Climate Change NSW, Sydney.
- 40. Meek P, Nazer C (1995) Vertebrate pests and weeds In: Cho G, Georges G, Stoutjesdijk R editors Jervis Bay. A Place of Cultural, Scientific and Educational Value. Kowari 5, Australian Nature Conservation Agency, Canberra. 115–122.
- 41. Vesk PE, Leishman MR, Westoby M (2004) Simple traits do not predict grazing response in Australian dry shrublands and woodlands. J Appl Ecol 41: 22–31.
- 42. Lange RT, Graham CR (1983) Rabbits and the failure of regeneration in Australian arid zone Acacia. Aust J Ecol 8: 377–381.
- 43. Cowling RM, Lamont BB (1987) Post-fire recruitment of four co-occurring Banksia species. J Appl Ecol 24: 645–658.
- 44. Cooke BD (1987) The effects of rabbit grazing on regeneration of sheoaks, Allocasuarina verticilliata and saltwater ti-trees, Melaleuca halmaturorum, in the Coorong National Park, South Australia. Aust J Eco 13: 11–20.
- 45. Semple WS, Koen TB (2001) Growth rate and effect of sheep browsing on young eucalypts in an anthropogenic Themeda grassland. Rangeland J 23: 182–193.
- 46. Scurr GF, Kirkpatrick JB, Daniels G, McQuillan PB (2008) Biotic resistance to Chrysanthemoides monilifera ssp. monilifera in Tasmania. Austral Ecol 33: 941–950.
- 47. Robinson K, Bell LW, Bennett RG, Henry DA, Tibbett M, et al. (2007) Perennial legumes native to Australia - a preliminary investigation of nutritive value and response to cutting. Aust J Exp Agr 47: 170–176.
- 48. Fensham RJ, Holman JE, Cox MJ (1999) Plant species responses along a grazing disturbance gradient in Australian grassland. J Veg Sci 10: 77–86.
- 49. Symon DE (1986) A survey of Solanum prickles and marsupial herbivory in Australia. Ann Missouri Bot Gard 73: 745–754.
- 50. Sprent J, McArthur C (2002) Diet and diet selection of two species in the macropodid browser–grazer continuum – do they eat what they ‘should’? Aust J Zool 50: 183–192.
- 51. Cooper-Driver G, Finch S, Swain T, Bernays E (1977) Seasonal variation in secondary plant compounds in relation to the palatability of Pteridium aquilinum. Bioch Syst and Ecol 5: 177–183.
- 52. Fenwick GR (1988) Bracken (Pteridium aquilinum) - toxic effects toxic constituents. J Sci Food Agr 46: 147–173.
- 53. Falvey JL (1981) Imperata cylindrica and animal production in south-east Asia: A review. Trop Grasslands 15: 52–56.
- 54. Edmonds JM, Chweya JA (1997). Black nightshades. Solanum nigrum L. and related species. Promoting the conservation and use of underutilized and neglected crops. 15. Institute of Plant Genetics and Crop Plant Research, Gatersleben International Plant Genetic Resources Institute, Rome, Italy.
- 55. Vranjic JA, Groves RH (1996) Revegetation of coastal habitats invaded by Bitou bush. In: Bellairs SM Osborne JM editors. Second Australian Native Seed Biology for Revegetation Workshop – Proceedings, ACT, 221–224.
- 56. Thomas J, Hofmeyer D, Benwell AS (2006) Bitou Bush control (after fire) in Bundjalung National Park on the New South Wales North Coast. Ecological Management and Restoration 7: 80–92.
- 57. Prattis KE (2004) An investigation into the effects of Chrysanthemoides monilifera ssp. rotundata (Bitou Bush) infestation and treatment on the vegetation community at Booderee National Park. Honours Thesis, University of Wollongong.
- 58. Booderee National Park Management Plan (2002), Consisting of a description of Booderee National Park and a management plan, Commonwealth of Australia.
- 59. Dexter N, Meek P (1998) An analysis of bait-takes and non-target impacts during a fox control exercise. Wildl Res 25: 147–155.
- 60. Von Ende C (2001) Repeated measures analysis: growth and other time dependent measures. In: Scheiner SM Gurevitch J editors. Design and Analysis of Ecological Experiments. Oxford University Press, Oxford, UK, 134–157.
- 61. McCulloch CE Neuhaus JM (2005) Generalized Linear Mixed Models. In: Armitage P, Colton T, editors, Encyclopedia of Biostatistics , John Wiley and Son, Chichester, 2085–2089.
- 62. Clarke KR, Gorley RN (2006) PRIMER v6 User Manual/Tutorial PRIMER-E: Plymouth.
- 63. Dexter N, Ramsey D, Macgregor C, Lindenmayer D (2012) Predicting ecosystem wide impacts of wallaby management using a fuzzy cognitive map. Ecosystems 15: 1363–1379.
- 64. Bergstrom DM, Lucieer A, Kiefer K, Wasley J, Belbin L (2009) Indirect effects of invasive species removal devastate World Heritage Island. J Appl Ecol 46: 73–81.
- 65. Robertshaw JD, Harden RH (1986) The ecology of the dingo in north-eastern New SouthWales. IV. Prey selection by dingoes and its effect on the major prey species, the swamp wallaby, Wallabia bicolor (Desmarest). Aust Wildl Res 13: 141–63.
- 66. Gill RMA (1992) A review of damage by mammals in north temperate forests. 3. Impacts on trees and forests. Forestry 65: 363–388.
- 67. Florence RG (1996) Ecology and silviculture of eucalypt forests. CSIRO, Collingwood.
- 68. McArthur C, Goodwin A, Turner S (2000) Preferences, selection and damage to seedlings under changing availability by two marsupial herbivores. Forest Ecol Manage 139: 157–173.
- 69. Vourc'h GB, Vila D, Gillon J, Escarre F, Guibal H, et al. (2002) Disentangling the causes of damage variation by deer browsing on young Thuja plicata. Oikos 98: 271–283.
- 70. McCaw WL (1988) Regeneration of Acacia and Kennedia from soil-stored seed following an autumn fire in jarrah (Eucalyptus marginata) forest. Journal of the Royal Society of Western Australia 71: 1–6.
- 71. Diaz S, Noy-Meir I, Cabido M (2001) Can grazing response of herbaceous plants be predicted from simple vegetative traits? J Appl Ecol 38: 497–508.
- 72. Niu K, Zhang S, Zhao B, Du G (2010) Linking grazing response of species abundance to functional traits in the Tibetan alpine meadow. Plant Soil 330: 215–223.
- 73. Hollis CJ, Robertshaw JD, Harden RH (1986) Ecology of the swamp wallaby (Wallabia bicolor) in north-eastern NSW. I. Diet. Aust Wildl Res 13: 355–361.
- 74. Hanley ME, Lamont BB (2001) Herbivory, serotiny and seedling defence in Western Australian Proteaceae. Oecologia 126: 409–417.
- 75. Rafferty C, Lamont BB, Hanley ME (2005) Selective feeding by kangaroos (Macropus fuliginosus) on seedlings of Hakea species: Effects of chemical and physical defences. Plant Ecology 177: 201–208.
- 76. Ens EJ, French K (2008) Exotic woody invader limits the recruitment of three indigenous plant species. Biol Cons 141: 590–595.
- 77. Rooney TP (2001) Deer impacts on forest ecosystems: a North American perspective. Forestry 74: 201–208.
- 78. Tolhurst KG, Turvey ND (1992) Effects of bracken (Pteridium esculentum (Forst. f.) Cockayne) on eucalypt regeneration in west-central Victoria. For Ecol Manage 54: 45–67.
- 79. Horsley SB, Marquis DA (1983) Interference by weeds and deer with Allegheny hardwood reproduction. Can J For Res 13: 61–69.
- 80. Hobbs NT (1996) Modification of ecosystems by ungulates. J Wildl Manage 60: 695–713.