South African Journal of Botany 87 (2013) 4–8

Contents lists available at SciVerse ScienceDirect

South African Journal of Botany

journal homepage: www.elsevier.com/locate/sajb

Short communication Effectiveness of guards in reducing grazing of Tetraria capillaris in restored bauxite mines in

Emily M. Stanton-Clements a,b, John M. Koch c, Matthew I. Daws c,⁎ a Newmont Asia Pacific, 388 Hay St., Subiaco, WA 6008, Australia b In the Field Environmental Consultancy, P.O. Box 149, Mt. Helena, WA 6082, Australia c Environmental Research Department, Alcoa World Alumina Australia, Huntly Mine, P.O. Box 172, Pinjarra, WA 6208, Australia article info abstract

Article history: Herbivores can exert significant impacts on vegetation composition, particularly in open, newly established Received 27 February 2013 restored sites where are visible and often represent high quality forage. While tree species can poten- Accepted 7 March 2013 tially outgrow herbivores, understorey plants are unable to do so and may remain prone to herbivory in de- Available online 9 April 2013 veloping restored sites. We assessed the effectiveness of various mechanical and biological guards ( companion plants) against grazing of the palatable dryland sedge Tetraria capillaris (hair sedge), in sites re- Edited by J Van Staden stored after bauxite mining in the jarrah forest of Western Australia. Six months after planting, grazing was Keywords: severe and repetitive, and affected plant survival. Grazing intensity and severity reduced over time (from 95% Herbivory to 64% of surviving plants grazed after six and a half years). No method of guarding entirely prevented graz- Kangaroo ing. Plants with plastic sleeves and Polynet mesh tubes showed significantly higher survival rates in short and Mine rehabilitation longer term compared with unprotected controls (62% survival for both treatments compared with 30% sur- Plant establishment vival of controls after six and a half years). Plastic sleeves were most effective at guarding T.capillaris against Restoration grazing although they were less cost-effective than Polynet tubes. While companion planting also reduced grazing it was less effective than artificial guards and not cost-effective. These data indicate that plant guards can have a positive impact on plant performance, even 6.5 years after planting. © 2013 SAAB. Published by Elsevier B.V. All rights reserved.

1. Introduction apparent six years after planting. While tree species have the potential to grow beyond the reach of grazers and thereby outgrow the usefulness Herbivores can have significant impacts on plant community com- of plant guards, the long-term benefits of plant guards on understorey position by preferentially grazing particular plant species (Agrawal, plants that remain at a ‘grazeable’ height into the long-term are unknown. 1998; Koch et al., 2004). In disturbed ecosystems, such as mine resto- Various studies have shown that in natural environments, establishing ration, herbivores can move unobstructed through large open areas seedlings can benefit from being in close proximity to ‘nurse’ plants containing highly visible plants/seedlings which are often nutrient (Franco and Nobel, 1989; Tewksbury and Lloyd, 2001) with this prin- enriched (Koch et al., 2004). Since the establishing vegetation in these ciple being applied in restoration efforts (Aerts et al., 2007; Castro et environments is all of the same, vulnerable, non-reproductive age, al., 2004). Growing within, or beneath, a nurse plant can provide pro- selective removal of seedlings has the potential to reduce or eliminate tection from grazing by providing physical (e.g. spines) or chemical a species from an area. (e.g. toxins) protection or simply reducing the visibility of establishing In Australia, grazing by kangaroos and other macropods has a plants to herbivores (Smit et al., 2007). However, the beneficial effects strong impact on the establishment of native plants in a wide variety of nurse plants are not always observed in restoration studies (Close of habitats (Abbott, 1984; Bell et al., 1987; Curtis, 1998; De Munk, 2001; et al., 2010) with increased competition resulting from the close proxim- Gardiner, 1986; Parsons et al., 2006, 2007). For establishing tree species, ity of neighbours potentially outweighing facilitation effects (Gómez- artificial plant guards have been trialed to reduce macropod grazing. For Aparicio et al., 2004). example, Close et al. (2010) found that for two Eucalyptus species, there Tetraria capillaris (Cyperaceae) is a perennial rhizomatous herb and was a site dependent effect of plant guards on growth and survival one of the most abundant species in jarrah forest being found at densi- that was related to grazing pressure. At the site with the higher grazing ties of 18,000–27,000 plants per hectare (Koch, 2007a). T. capillaris is pressure, the tree guard effect on height growth and survival was still also a favourite food source for kangaroos and can be killed by intensive grazing pressure (Parsons et al., 2006). Following bauxite mining and restoration, this species establishes at very low densities (≤25 ha−1; ⁎ Corresponding author. Tel.: +61 8 95302340; fax: +61 8 95302571. Alcoa, unpublished data). Given that T. capillaris produces very limited E-mail address: [email protected] (M.I. Daws). quantities of viable seed, it is grown by tissue culture and planted into

0254-6299/$ – see front matter © 2013 SAAB. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.sajb.2013.03.006 E.M. Stanton-Clements et al. / South African Journal of Botany 87 (2013) 4–8 5 restored mine pits to increase the density in restored areas (Willyams, 4. Acacia lateriticola seedling (non-thorny Acacia) planted adjacent to 2005). each T. capillaris; The experiment described here was established to determine the 5. Acacia pulchella seedling (thorny Acacia) planted adjacent to each extent and impact of grazing by western grey kangaroos (Macropus T. capillaris; fuliginosus) and black gloved wallabies (Macropus irma)onT.capillaris. 6. A. lateriticola seeds (×3) planted adjacent to each T. capillaris; The study also examined the effectiveness of several types of artificial 7. A.pulchella seeds (×3) planted adjacent to each T. capillaris; and guards and biologically guarding T.capillaris against grazing by herbi- 8. Control (no treatment). vores. This was achieved using a range of commercial guards and native Acacia species as nurse plants. Many studies of herbivore impacts on The plastic sleeves (1) were provided by Nurserymen's Supplies restoration only examine early establishment stages. However, since (Kewdale, Perth, Australia). The transparent, green sleeves (350 × T. capillaris does not grow beyond the reach of grazers, the experiment 450 mm) were made of a thin (about 250 μm), but relatively tough was continued for six and a half years after establishment to look at layer of plastic, and punctured by 8 holes that allowed some airflow longer-term impacts on plant performance. The experiment had the across them. The plastic sleeves were erected around the seedling following aims: at planting, by placing them over two jarrah stakes (750 × 20 mm) which were hammered into the ground. The sleeve was then stretched 1. Determine the survival and growth of T.capillaris for up to 6.5 years out and stabilised by the installation of a third stake. The sleeve was after planting in restored jarrah forest; stapled at two points (top of one stake and bottom of another) to pre- 2. Determine whether grazing of T.capillaris continues to be high in vent the sleeve from slipping. restored jarrah forest or declines after planting; and The plastic mesh guards (2) were obtained from Nurserymen's 3. Determine the ongoing effectiveness of various methods for guarding Supplies (Kewdale, Perth). The guards were 230 mm in diameter T.capillaris against herbivory in restored jarrah forest. and 450 mm high and made of clear, rigid plastic mesh, composed of 30 × 30 mm diamonds. The mesh guards were placed over the 2. Methods T. capillaris seedlings and held in place by 2 bamboo sticks (750 × 10 mm) that were woven through the top, middle and bottom of 2.1. Site description the mesh and then pushed by hand into the ground. Half of a third bamboo stick was then lodged through one side of the guard into The experiment was established in August 2001, at the Huntly the ground to act as an anchor against the wind. (Alcoa of Australia Ltd, 32°35′06″ S, 116°06′44″ E) and Boddington The Polynet tubes (3) were made of a fine, green, plastic mesh (BHP Billiton Worsley Alumina Pty Ltd, 32°53′36″ S, 116°27′12″ E) (Koch, 2007b), which is produced on a compact roll by Sarlon (Perth, bauxite mines in the southwest jarrah forest of Western Australia. Australia). The dimensions of the diamond mesh on the tube were The two mines are in the Darling Range, southwestern Australia ca. 5 × 5 mm. The net was cut as required (to about 300 mm length) 100–180 km south of Perth. The region has a Mediterranean-type cli- and, prior to planting, pulled over the plant and Jiffy pot (Fertil S.A., mate with cool, wet winters and hot, dry summers. Average annual Le Syndicat, France) by hand. The mesh tube formed a free-standing rainfall is 1200 mm p.a. at Huntly and 710 mm at Boddington. Average guard, which required no additional support mechanism. summer monthly maximum temperatures are 28 °C at Huntly and Acacia seedlings (4–5) were grown in 50 × 50 mm Jiffy pots from 30 °C at Boddington. Average winter minima are 5 °C at both sites. seeds sown in April 2001. At the time of planting (August 2001), the The sclerophyll vegetation within the mining envelopes is generally seedlings were 4 months old, with a mean height of 50 mm. One dominated by trees of jarrah () with varying den- seedling was planted 100 mm distant from each T. capillaris plant. sities of marri (). In addition, there is a small-tree The A. lateriticola and A.pulchella seeds (6–7) were boiled for 1 min component, with Banksia grandis, Allocasuarina fraseriana, Persoonia to promote germination. One seed (A. lateriticola or A. pulchella)was longifolia and Xylomelum occidentale being the most common species. placed at 3 points, 100 mm from the T. capillaris seedling and about The undergrowth consists of sclerophyllous shrubs (to 3 m tall) and 5 mm under the soil surface. herbs predominantly in the families Myrtaceae, Proteaceae, , T.capillaris plants were grown by tissue culture at Alcoa's Marrinup Restionaceae, Orchidaceae, Apiaceae, Liliaceae (sensu lato), Ericaceae, Nursery (detailed methods described by Willyams, 2005). At the time Asteraceae and Cyperaceae. of planting, the T. capillaris plants were 7 months old and growing in The mine pits range in size from 1 to 20 ha and the restoration 50 × 50 mm Jiffy pots. The plants had a mean height of 87.8 ± 24.7 process involves deep ripping the pit floor to alleviate compaction, (sd) mm. T. capillaris and Acacia seedlings were planted in August 2001. returning the topsoil followed by seeding with a diverse mix of local Plants were assessed twice: in February 2002 and between December plants. The overall aim of the restoration is to recreate a jarrah forest 2007 and February 2008. During both rounds of monitoring, T.capillaris ecosystem. Planting of nursery grown plants is carried out for species survival, maximum height and occurrence of grazing were determined. where broadcast seed is not a viable establishment method: these are Additionally, in 2007/08, above-ground biomass, plant spread (maximum grown from seeds, cuttings or tissue culture. distance between shoots) and percentage of plants that were flowering/ seeding were recorded. Above-ground biomass was determined by cut- 2.2. Experimental design ting the foliage to ground level and then determining the dry mass after the clippings had been dried in an oven at 65 °C for 48 h. Guard place- At each mine, three replicate pits were chosen, all of which were ment and persistence at 6.5 years were also recorded. in previously mined areas. These sites were restored and seeded in 2001 following normal restoration processes. Within each pit, 160 in- 2.3. Statistical analysis dividual T.capillaris were planted in a grid ca. 1 m apart, with each of the eight guard treatments applied to 20 randomly allocated plants. One-way ANOVA, followed by Tukey's post hoc test, implemented The guard treatments were: in Minitab 16 was used to test for effects of guarding treatment, on survival, incidence of grazing, plant height, biomass and spread. The 1. Plastic sleeve stapled to three wooden (jarrah) stakes; performance of each plant guard in relation to plant traits in 2007/08 2. Rigid plastic mesh sleeve with two interwoven bamboo stakes and was characterised using survival (percentage), grazing (percentage of an additional half stake to anchor the mesh to the ground; plants), above ground biomass (g), plant spread (mm) and proportion 3. Polynet tube (self supporting); of plants flowering. The 8 × 5 guard treatment × plant response matrix 6 E.M. Stanton-Clements et al. / South African Journal of Botany 87 (2013) 4–8 was submitted to principal component analysis (PCA) to explore the A 100 intercorrelation amongst plant responses to the guard treatments. 2002 a a 2007/08 3. Results 80 ab A b ab A b b b 3.1. Occurrence of fauna 60 AB ABC ABC In both 2002 and 2007/08, all sites showed evidence of visitation BC BC (including scats, tracks and resting places) by western grey kangaroos Survival (%) 40 C and black gloved wallabies, with the exception of Boddington site 1, which had no sign of wallabies. 20 3.2. Persistence of plant guards 0 Plastic sleeves were no longer in place in 2008 with remains of B a a a a a a 100 a A disintegrated sleeves observed occasionally amongst the litter. However, 99% of the wooden stakes of this treatment were still in A b 80 A situ. While the plastic of the Polynet tubes and plastic mesh guards A AB AB appeared not to have disintegrated, 23 and 41% of these two treat- AB ments, respectively, were not in place in 2008 because they had be- 60 B come dislodged or fallen over. For the biological guards, A. pulchella from both seedlings and seed was established to higher levels than 40 A. lateriticola. Thus, 67 and 59% of A. pulchella guards from seedlings and seeds, respectively, were in place in 2008 compared with 50 Individuals grazed (%) and 37% of A. lateriticola guards from seedlings and seed, respectively. 20 By 2008, a number of Acacia guards from both seed and seedlings had senesced at maturity. These were recorded as being in place since 0 they still provided physical defence to the T. capillaris plants. PolynetPlastic Mesh Ap Al Ap Al Control At the time of the initial 2002 assessment, successful establish- tube sleeve s'ling s'ling seed seed ment of Acacia guards from seeds was quite low, with 30 and 41% of Fig. 1. The effect of different plant guard treatments on (A) survival and (B) the per- A. lateriticola and 31 and 24% of A. pulchella seeds successfully established centage of individuals showing evidence of being grazed for Tetraria capillaris planted at Huntly and Boddington, respectively. Growth of from seed into restored sites in 2001. Survival and the occurrence of grazing were monitored in was slow and at the time of the 2002 assessment, A. lateriticola had both 2002 and 2007/08. Within each plant response, monitoring year combination, mean heights of 50 and 65 mm at Huntly and Boddington respectively. letters adjacent to symbols indicate significant differences (P b 0.05) within each plant A. pulchella had slightly greater mean heights of 68 (Huntly) and response in relation to guard treatment. Ap = Acacia pulchella;Al=Acacia lateriticola; s'ling = planted seedling; seed = planted seed. 129 mm (Boddington). The small stature of acacias from seeds meant that neither species covered T. capillaris seedlings at the initial monitor- ing interval, offering no physical protection from herbivory. In contrast to the Acacia guards from seeds, survival of Acacia guards exhibited some evidence of grazing). Similar to 2002 results, there from seedlings in 2002 was relatively high, with 85% of A. lateriticola was a significant effect of guarding treatment on grazing incidence, (at both Huntly and Boddington) and 75 (Huntly) and 82% (Boddington) and plants assigned the plastic mesh treatment had the lowest inci- of A. pulchella surviving. A. lateriticola reached a mean height of 318 mm dence of grazing (one-way ANOVA, P b 0.001; Fig. 1B). while A. pulchella reached 366 mm. A. pulchella developed thorns about Plant spread over the course of the 6.5 years varied significantly two months after planting and soon became prickly to touch. Of the between treatments (one-way ANOVA, P b 0.001; Fig. 2A) ranging T. capillaris plants assigned A. lateriticola and A. pulchella seedlings, 10% from ca. 25 mm in the control to ca. 100 mm in the plants assigned and 20% were covered by Acacia foliage, respectively. At the 2007/08 a Polynet bag. Plant spread decreased in the general order: plants assessment, Acacia growth from seed and seedlings was similar reaching with artificial plant guards > those assigned Acacia seedlings > those approximately 900 and 1800 mm for A. lateriticola and A. pulchella, assigned Acacia seeds > the control. Plant height in 2007/08 was also respectively. significantly related to guard treatment (one-way ANOVA, P b 0.001; Fig. 2A) and followed a similar pattern to spread: plants assigned artifi- 3.3. Effect of time and plant guards on herbivory, survival and growth of cial plant guards were ca. 3 times taller than those in the control. Plant T. capillaris height also decreased in the same general order: plants with artificial plant guards > those assigned Acacia seedlings > those assigned Acacia At both assessments (2002 and 2007/8), survival was only signif- seeds > the control. icantly higher than the control (P b 0.05; Fig. 1A) for the Polynet bag The proportion of individuals flowering/seeding in 2007/08 was and plastic sleeve guards. Survival declined in all treatments over also significantly related to guarding treatment (one-way ANOVA, time, reducing from ca. 82% in 2002 to ca. 60% in 2007/8 in the Polynet P b 0.001; Fig. 2B) and followed the same general decline as observed bag and plastic sleeve guards, and from ca. 65% to ca. 30% for the control with both plant spread and height: ca. 25% of plants with artificial (Fig. 1A). guards flowered compared with ca. 5% in the control. Furthermore, There was widespread occurrence of grazing on the T. capillaris the biomass at the end of the experiment was significantly related to plants at all replicate sites in both 2002 and 2007/08. In 2002, 75 to guard treatment with biomass spanning from ca. 5 g (plastic sleeves) 98% of plants (dependent on treatment) showed some evidence of to 0.8 g (control) (one-way ANOVA, P b 0.001; Fig. 2B). Plants assigned grazing. The incidence of grazing in 2002 was significantly lower for artificial plant guards had a higher biomass than those assigned Acacia plants with a plastic sleeve than the other six treatments and the con- seeds and the control although those assigned Acacia seedlings had trol (one-way ANOVA, P b 0.05; Fig. 1B). In 2007/08, the overall inci- a similar biomass to those assigned Polynet bags and plastic mesh dence of grazing had decreased (across treatments, 64% of plants (Fig. 2B). E.M. Stanton-Clements et al. / South African Journal of Botany 87 (2013) 4–8 7

400 A Spread a Height

300 ab ab ab ac 200

bc bc Spread / height (mm) a c 100 a a ab ab bc bc c 0 40 B Flowering 20 Above-ground biomass

30 15 a a

ab 20 ab 10 ab Biomass (g) ab a b 5 10

Individuals flowering / fruiting (%) ab ab ab ab b bc c c 0 0

PolynetPlastic Mesh Ap Al Ap Al Control tube sleeve s'ling s'ling seed seed

Fig. 2. The effect of different plant guard treatments on (A) both plant spread and height and (B) the percentage of individuals flowering or seeding and above-ground biomass for Tetraria capillaris planted into restored sites in 2001 and monitored in 2007/08. Letters adjacent to symbols indicate significant differences (P b 0.05) within each plant response in relation to guard treatments. Ap = Acacia pulchella;Al=Acacia lateriticola; s'ling = planted seedling; seed = planted seed.

3.4. Principal component analysis artificial plant guards, and to some extent nurse plants, can reduce the incidence and impact of grazing, with these effects still evident The PCA axis I had an eigenvalue of 3 · 75, and explained 74.9% of 6.5 years after planting. the variance in the 2007/08 plant traits. The eight guard treatments The PCA analysis indicated that T. capillaris plants exhibited a suite were arranged along this axis as follows with axis I scores: plastic of inter-related responses to grazing with the different guard treat- sleeve (−2.39), onion bag (−2.16), mesh (−1.23), A. lateriticola ments separating along PCA axis I, such that plants assigned the plastic seedling (−0.32), A. pulchella seedling (−0.31), A. lateriticola seed sleeves had the lowest grazing impacts and the control the most severe (+1.81), A. pulchella seed (+1.87) and control (+2.71). All pairwise impacts. Thus, artificial plant guards, followed by planting Acacia seed- correlations between the plant attributes (biomass, survival, spread lings and Acacia seeds increased survival, spread, flowering and above- and percentage of individuals flowering) were significant (P b 0.05). ground biomass. While there was only a limited effect of guard type However, none of the pair-wise correlations involving the percentage on the percentage of plants grazed, this was presumably because of plants that were grazed were significant (P > 0.05). In addition, all guards decreased the extent of, but did not entirely prevent, grazing. traits except for the percentage of plants grazed were significantly Although the proportion of plants grazed decreased generally over- correlated with PCA axis I. time (including for the control plants), our results indicate that plants exposed to high initial grazing under-perform over the medium term 4. Discussion compared with individuals protected from grazing (by artificial guards) from the outset. Grazing pressure likely decreased over time due to a The data indicate that within seven months of planting, T. capillaris transition from open and easily traversable sites to increasing habitat plants experienced high grazing pressure and an impact of guard type complexity associated with tree and understorey growth and litter ac- on survival. In addition, even 6.5 years after planting, T. capillaris still cumulation providing a general concealing effect for the T. capillaris displayed an effect of grazing with impacts on survival, plant spread, plants (i.e. the plants were less apparent; c.f. Castagneyrol et al., flowering and biomass accumulation. These results demonstrate that 2013). 8 E.M. Stanton-Clements et al. / South African Journal of Botany 87 (2013) 4–8

The three artificial plant guards were the most effective treat- with plants spreading laterally and flowering. In addition, monitoring ments, presumably because they were in situ and effective from the over this time-frame indicates that there is an enduring effect of plant time of planting, and had a generally high level of persistence over guards on plant performance and that plant guards can be cost-effective time. In contrast, the Acacia plant guards (particularly from seed) even when used in large-scale restoration. Overall these data indicate took time to develop and provide protection to the T. capillaris plants, that these plants are most likely on a trajectory towards producing a during which plants were exposed to grazing. Nonetheless, compared self-sustaining population. with the control, the Acacia plant guards provided some level of pro- tection (particularly at 6.5 years) from herbivory indicating a positive Acknowledgements ‘nurse’ plant effect. Positive effects of nurse plants are generally only found in high stress environments, such as the highly seasonal Many thanks to Stephen Vlahos (BHP Billiton Worsley Alumina), Mediterranean climate of the jarrah forest and positive facultative Brad Stokes (BHP Billiton Worsley Alumina), Rory Haymont (Newmont), effects have been reported previously for jarrah forest species (Parsons Tim Morald (University of Western Australia), Melissa Graham (Univer- et al., 2007). sity of Western Australia) and Alice Bott (Curtin University) for provision In addition to decreasing grazing, plastic sleeves may change of advice, resources, field support and assistance with statistical analysis the micro-climate for plant growth. For example, Close et al. (2009) and editing. MID is a current paid employee of Alcoa of Australia Ltd examining seedling survival and growth of native tree species with while JMK is a former paid employee. different tree guards, in Perth Western Australia, showed that average temperature was 6.7 °C higher in plastic guards than controls over References summer, while light levels were twofold lower. At midday, photo- fi fi Abbott, I., 1984. Emergence, early survival, and growth of seedlings of six tree species chemical ef ciency was signi cantly higher compared with controls. in Mediterranean forest of Western Australia. Forest Ecology and Management 9, Close et al. (2009) found that the increase in temperature led to ele- 51–66. vated seedling mortality and suggested that plastic sleeves may be Aerts, R., Negussie, A., Maes, W., November, E., Hermy, M., Muys, B., 2007. Restoration of dry Afromontane forest using pioneer shrubs as nurse plants for Olea europaea inappropriate for use in Mediterranean-type ecosystems. However, ssp. cuspidata. Restoration Ecology 15, 129–138. in the Mediterranean-type ecosystem of the jarrah forest, we found Agrawal, A., 1998. Induced responses to herbivory and increased plant performance. no negative impact of plastic sleeves on survival. This absence of an Science 279, 1201–1202. effect may be because temperatures are more moderate in the jarrah Bell, D.T., Moredoundt, J.C., Loneragan, W.A., 1987. Grazing pressure by the tammar wallaby (Macropus eugenii Desm.) on the vegetation of Garden Island, Western forest than in Perth. Australia, and the potential impact on food reserves of a controlled burning Although the plastic sleeves were the most effective plant guard regime. Journal of the Royal Society of Western Australia 69, 89–94. treatment, they have the disadvantage that the plastic disintegrates Castagneyrol, B., Giffard, B., Péré, C., Jactel, H., 2013. Plant apparency, an overlooked driver of associational resistance to insect herbivory. Journal of Ecology. http:// after several years and can result in a litter nuisance. In addition, at dx.doi.org/10.1111/1365-2745.12055. the time of the initiation of the experiment, these guards were the Castro, J., Zamora, R., Hódar, J.A., Gómez, J.M., Gómez-Aparicio, L., 2004. Benefits most expensive treatment costing $1.53 AUD each. In comparison, of using shrubs as nurse plants for reforestation in Mediterranean mountains: a 4-year study. Restoration Ecology 12, 352–358. the mesh and Polynet guards cost $0.37 and $0.10 each, respectively, Close, D., Ruthrof, K., Turner, S., Rokich, D., Dixon, K., 2009. Ecophysiology of species and were nearly as effective. While they provided some level of benefit, with distinct leaf morphologies: effects of plastic and shadecloth tree guards. the Acacia seedlings cost $0.49 each which means that they were un- Restoration Ecology 17, 33–41. Close, D.C., Davidson, N.J., Churchill, K.C., Corkrey, R., 2010. Establishment of native economic compared with the more effective mesh and Polynet guards. Eucalyptus pauciflora and exotic Eucalyptus nitens on former grazing land. New While they were cheapest ($0.01 per guard), the cost saving associated Forests 40, 143–152. with using Acacia seeds was more than outweighed by the higher Curtis, N.P., 1998. A post-fire ecological study of Xanthorrhoea australis following prescribed burning in the Warby Range State Park, northern-eastern Victoria, mortality of the T. capillaris plants compared with using Polynet bags Australia. Australian Journal of Botany 46, 253–272. (survival was ca. 35% lower): the production cost of the T. capillaris De Munk, F.G., 2001. Resource use by the eastern grey kangaroo and the black wallaby plants in 2004 was $3.93 each (Willyams, 2005). Overall, the Polynet in a managed remnant woodland community. Australian Biologist 14, 11–12. tubes were the most cost-effective guard type for T. capillaris (Table 1). Drayton, B., Primack, R.B., 2012. Success rates for reintroductions of eight perennial plant species after 15 years. Restoration Ecology 20, 299–303. Drayton and Primack (2012) highlighted that studies of the success Franco, A.C., Nobel, P.S., 1989. Effects of nurse plants on the microhabitat and growth of of plant establishment in restoration, have only generally assessed cacti. Journal of Ecology 77, 870–886. initial or early-establishment, while to fully understand population es- Gardiner, H.G., 1986. Dynamics of perennial plants in the Mulga (Acacia aneura F. Muell.) zone of Western Australia 11. Survival of perennial shrubs and grasses. tablishment, longer monitoring time frames are needed. In our current The Australian Rangeland Journal 8, 28–35. study, survival of planted T. capillaris was monitored over a 6.5 year Gómez-Aparicio, L., Zamora, R., Gómez, J.M., Hódar, J.A., Castro, J., Baraza, E., 2004. Ap- time-frame and was ca. 60% in the most effective guarding treatments plying plant facilitation to forest restoration: a meta-analysis of the use of shrubs as nurse plants. Ecological Applications 14, 1128–1138. Koch, J.M., 2007a. Restoring a jarrah forest understorey vegetation after bauxite mining in Western Australia. Restoration Ecology 15, S26–S39. Table 1 Koch, J.M., 2007b. Alcoa's mining and restoration process in South Western Australia. Relative costs of the different plant guard treatments. Also shown for each guard treat- Restoration Ecology 15, S11–S16. ment is survival of Tetraria capillaris plants to 6.5 years after planting and the resulting Koch, J.M., Richardson, J., Lamont, B.B., 2004. Grazing by kangaroos limits the establish- ‘cost’ of each surviving plant at 6.5 years. ment of the grass trees Xanthorrhoea gracilis and X.preissii in restored bauxite mines in eucalypt forest of southwestern Australia. Restoration Ecology 12, 297–305. Guard treatment Guard cost Tetraria capillaris Cost per surviving Parsons, M.H., Koch, J.M., Lamont, B.B., Vlahos, S., Fairbanks, M.M., 2006. Planting density ($AUD per stem)a survival (%) plant ($AUD)b effects and selective herbivory by kangaroos on species used in restoring forest com- munities. Forest Ecology and Management 229, 37–49. Polynet tube 0.10 62.5 6.45 Parsons, M.H., Lamont, B.B., Koch, J.M., Dods, K., 2007. Disentangling competition, Plastic sleeve 1.53 62.3 8.77 herbivory and seasonal effects on young plants in newly restored communities. Plastic mesh 0.37 49.3 8.72 Restoration Ecology 15, 250–262. Acacia pulchella 0.49 44.1 10.02 Smit, C., Vandenberghe, C., Den Ouden, J., Müller-Schärer, H., 2007. Nurse plants, tree seedling saplings and grazing pressure: changes in facilitation along a biotic environmental A. lateriticola seedling 0.49 46.1 9.59 gradient. Oecologia 152, 265–273. A. pulchella seed 0.01 29.1 13.51 Tewksbury, J.J., Lloyd, J.D., 2001. Positive interactions under nurse-plants: spatial scale, A. lateriticola seed 0.01 34.5 11.42 stress gradients and benefactor size. Oecologia 127, 425–434. Control (unguarded) N/A 30.5 12.89 Willyams, D., 2005. Tissue culture of geophytic rush and sedge species for revegetation of bauxite mine sites in the northern jarrah forest of Western Australia. In: Bennett, a Based on the cost in 2001. I.J., Bunn, E., Clarke, H., McComb, J.A. (Eds.), Contributing to a Sustainable Future. b Based on 2001 plant guard cost and 2004 production costs of $3.93 per plant Proceedings of the Australian Branch of the IAPTC&B, Perth, Western Australia, (Willyams, 2005). pp. 226–241.