RESEARCH doi: 10.1111/j.1442-8903.2011.00568.x REPORT Planted saltbush (Atriplex nummularia)andits value for birds in farming landscapes of the South Australian Murray Mallee

By Stuart J. Collard, Andrew M. Fisher and David J. McKenna

Stuart Collard is Senior Environmental Consultant, Summary In the fragmented agricultural landscapes of temperate southern Australia, Rural Solutions (GPO Box 1671, Adelaide, SA 5001, broad-scale revegetation is underway to address multiple natural resource management Australia; Tel: +61 8 8463 7885; Fax: 61 8 8301 issues. In particular, commercially-driven fodder shrub plantings are increasingly being 5720; Email: [email protected]). Andrew established on non-saline land to fill the summer-autumn feed gap in grazing systems. Little Fisher is Principal Advisor, Landscape Management, is known of the contribution that these and other planted woody perennial systems make to SA Department of Environment and Natural biodiversity conservation in multifunctional landscapes. In order to address this knowledge Resources and Future Farm Industries Co-operative gap, a study was conducted in the southern Murray Mallee region of South Australia. Research Centre (GPO Box 1047, Adelaide, SA 5001, Selected ecological indicators, including plant and bird communities, were sampled in spring Australia; Email: andrew.fi[email protected]). David 2008 and autumn 2009 in five planted saltbush sites and nearby areas of remnant vegetation McKenna was a Research Scientist when this work and improved pasture. In general, remnant vegetation sites had higher biodiversity values was undertaken but is now a Fire Management than saltbush and pasture sites. Saltbush sites contained a diverse range of plants and birds, Officer with the SA Department of Environment and including a number of threatened bird species not found in adjacent pasture sites. Plant and Natural Resources (GPO Box 1047, Adelaide, SA bird communities showed significant variation across saltbush, pasture and remnant treat- 5001, Australia; Email: [email protected]). ments and significant differences between seasons. This study demonstrates that saltbush This research was conducted as part of a broader plantings can provide at least partial habitat for some native biota within a highly modified Department of Environment and Natural Resources agricultural landscape. Further research is being conducted on the way in which biota, such project on the opportunities for establishing planted as birds, use available resources in these dynamic ecosystems. An examination of the effects woody perennial vegetation in the agricultural land- of grazing on biodiversity in saltbush would improve the ability of landholders and regional scapes of South Australia. natural resource management agencies in making informed land management decisions. Key words: agricultural landscape, Atriplex, biodiversity, birds, farm management, fodder shrubs, saltbush.

environments for native species (Attwood Fodder shrub plantings have the poten- Introduction et al. 2009; Prober & Smith 2009). tial to enhance landscape-scale heterogene- gricultural intensification across the In marginal farming areas of southern ity and local-scale habitat structural Atemperate landscapes of southern Aus- Australia, an increasing number of land diversity, thus enhancing biodiversity con- tralia has resulted in the well-documented managers are using alternative manage- servation efforts and supplementing exist- fragmentation and extensive degradation ment systems, including perennial fodder ing perennial plantings and stands of of natural vegetation systems, leading to shrubs such as saltbush to broaden the native vegetation (Collard & Fisher 2010). biodiversity loss at local and landscape feed base over summer and autumn in Across South Australia, over 7000 ha of scales (Hobbs 1993). Efforts to restore hab- grazing systems (McKenna et al. 2009). saltbush (mostly Old Man Saltbush, Atri- itat and ecosystem function within these The production aspects of growing shrub- plex nummularia) was planted between landscapes often fall short of achieving the based systems such as saltbush in low to 1999 and 2008 (DWLBC 2008). In the Mur- level that is required to meet regional res- moderate rainfall areas are well docu- ray Mallee region of South Australia, an toration and land management targets. The mented (Bartel & Knight 2000; McKenna estimated 7120 ha of woody revegetation shortfall may be due to constraints associ- et al. 2009). However, such plantations was planted during the same period, of ated with costs, incompatibility with cur- also have the potential to play a role in which 1700 ha (24%) comprised saltbush rent (short rotation) farming practices, lack achieving multiple NRM objectives for bio- plantings. A short-term target of 4400 ha of direct economic benefits from habitat diversity conservation and sustainable land was set in 2001 for the establishment of restoration activities and inadequate incen- management (Lefroy & Smith 2004). In fodder blocks across the Murray Mallee in tives for farmers to change their land man- particular, Prober and Hobbs (2008) advo- the regional revegetation plan as part of agement practices (Morrison et al. 2008). cate a target of 30% of perennial produc- the Mallee Futures Program (MMLAP Limited uptake of revegetation by land- tion systems in degraded woodland 2001), with an overall target of 108 000 ha holders is paralleled by increasing recogni- landscapes, arguing that they can augment of newly established perennial vegetation tion by ecologists of the importance of resources for native species and help considered necessary to address a variety elements of these managed production restore viable farm incomes. of NRM issues in the region.

ª 2011 Ecological Society of Australia ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 37 RESEARCH REPORT

The increase in structural complexity with saltbush contractors and local farmers each of the replicate sites in spring 2008. resulting from the establishment of large also helped to identify potential sites. Quadrat locations were recorded using a areas of planted woody vegetation may Within the target area, fifteen study sites GPS and marked permanently so that they provide important resources for native (>2 ha) were selected, comprising five rep- could be relocated and surveyed again in biota when considered at a regional level licates of three different land management autumn 2009 to quantify temporal varia- (Munro et al. 2007; Smith 2009; Collard & ‘treatments’, namely: (i) saltbush planta- tion in vegetation attributes. Photo points Fisher 2010). Despite its growing presence tion – fodder blocks of Old Man saltbush were established at one corner of each veg- in the agricultural landscapes of temperate (A. nummularia) >3 years old, typically etation quadrat. The modified vegetation Australia, few published studies have planted in a regular grid pattern with indicators selected from Croft et al. (2005) assessed the value of planted saltbush for planted or volunteer native and exotic were: biodiversity. Exceptions are Norman et al. groundcover of grasses and herbs, grazed (2008) who examined plant and inverte- by sheep annually during late summer and Plant species richness brate assemblages in saltbush plantings on autumn, (ii) improved pasture – areas of salt-affected land in south-western Western fallow cropping land adjacent to sampled The plant species present in each quadrat Australia and Seddon et al. (2009) who saltbush patches, comprising a mix of (native and exotic) were recorded, with sampled bird species in planted saltbush native and exotic grasses and herbs, grazed voucher specimens collected for each spe- alleys in western New South Wales. by sheep year round, and (iii) remnant veg- cies for correct identification. Despite a number of published articles that etation – isolated patches of native mallee have identified the potential benefits of vegetation with largely intact canopy and Structural diversity (Ground cover) fodder shrub systems for biodiversity (e.g. shrub layers and light to moderate grazing Millsom 2002; Newton & Yunusa 2002; and disturbance of understorey and ground The percentage cover of leaf litter, Lefroy et al. 2005), few studies demon- cover vegetation by rabbits and infre- exposed rock, microphytic crust, native strate which species inhabit these plant- quently sheep. ground cover, exotic ground cover and ings or which resources they are utilising. bare ground was estimated. Sampling protocols This article presents findings from components of a preliminary investiga- Sampling was conducted to investigate dif- Structural diversity (Plant life forms) tion examining key indicators of biodiver- ferences in the structure and composition sity in Old Man Saltbush plantings. The of the vegetation across treatments and to The percentage cover of the different plant aim of the study was to improve under- record the bird species using these respec- life forms present in each quadrat was esti- standing of the biodiversity and resource tive treatments. mated. Life form attributes assessed values associated with planted saltbush The indicators sampled were selected included trees, shrubs, herbs, mat plants, compared with other landscape elements to represent aspects of biodiversity rele- grasses, tussocks, vines and climbers, mis- that exist along a gradient of structural vant to fragmented, multiple-use agricul- tletoe and ferns. complexity and agricultural management tural landscapes. Plants were selected as a Bird diversity and community intensification. We sought to establish measure of local-scale biodiversity and composition the structural and compositional features functional values, while birds were chosen of the saltbush vegetation, to compare as a measure of landscape-scale ecological All study sites were surveyed three times these features with those of other land processes and for their ability to use spa- in both the spring and autumn seasons for uses and to determine which faunal taxa tially and temporally variable resources. bird abundance and community composi- (birds) were associated with these differ- All sites were sampled for vegetation tion. Surveys were conducted in the morn- ent landscape elements. and birds across two seasons, spring (Octo- ing in fine conditions between 30 min and ber 2008) and autumn (March 2009), coin- 4 h after sunrise and consisted of a 2 ha, ciding with times of low and high grazing 20 min survey [based on the ‘Birds Austra- Methods intensity in saltbush. lia Atlas’ protocol of Barrett et al. (2003)]. Bird activity was recorded within a 2 ha Site selection Vegetation structure and area in the habitat of interest, with devia- community composition Study sites were located within a highly tions from a central transect to identify fragmented agricultural landscape near A modified form of the Bushland Condition birds. Species presence and abundance Pinnaroo (3515¢37.31¢¢ S, 14054¢30.47¢¢ Monitoring methodology devised for mal- were recorded, as well as behavioural E) in the southern Murray Mallee region of lee vegetation associations of the Southern observations where possible. During sam- South Australia. Potential sites were identi- Mount Lofty Ranges (Croft et al. 2005) was pling, birds that flew across the transect fied using a combination of spatial imagery used to assess vegetation characteristics. were considered in the analyses only if (aerial photos) and previously recorded Briefly, a representative 30 m by 30 m they were observed flying in close associa- locations of saltbush plantations from a quadrat was selected within each of the tion with the vegetation of the transect range of databases. Further consultation treatments (remnant, saltbush, pasture) at area.

38 ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 ª 2011 Ecological Society of Australia RESEARCH REPORT

for these differences were then identified In spring 2008, the type and relative Data analysis using the Similarity Percentage (SIMPER) amount of ground cover varied between Plant species richness data (native and exo- routine (Clarke & Gorley 2006). treatments (Table 1). Remnants had a tic species combined) and bird abundance higher proportion of microphytic crust and species richness data were compared and a lower proportion of introduced across seasons and treatments with two- Results ground cover compared with saltbush and way crossed analysis of variance (ANOVA) pasture treatments. Vegetation composition and using ‘Statistix’ version 8. Plant and bird In autumn 2009, the relative cover of structure species richness data were log (x +1) leaf litter increased in all treatments and transformed to address the distributional Overall, 89 plant species were recorded the amount of exotic cover in saltbush and and variance assumptions required for across all sites during spring and autumn pasture treatments diminished compared linear models. sampling combined (Appendix S1). Of the with cover in spring 2008 (Table 1). Micro- Analyses of the plant and bird commu- 82 species that were able to be identified, phytic crust was present only in remnant nity data (native and exotic species com- 35 were native and 47 were exotic sites in autumn. bined) were conducted using routines (Table 1). Purple Crassula (Crassula More plant life forms were present in within the multivariate statistical package peduncularis) is listed as ‘Rare’ in South remnant sites than saltbush or pasture in PRIMER v6 (Clarke & Gorley 2006). Plant Australia and was recorded in three of the both spring 2008 and autumn 2009 data were in presence ⁄ absence form and remnant sites in spring. Eighty-one species (Table 1). The reduction in total cover in required no transformation. Bird abun- were recorded across all sites in spring saltbush and pasture sites in autumn 2009 dance data were pooled across the three 2008 and 35 species were recorded in was largely due to a reduction in cover of repeat surveys within each season and 4th autumn 2009. A total of 58 species was herbs and low grasses. Tall shrubs made root transformed to reduce the influence recorded in remnant sites (39 were exclu- up the highest proportion of plant life form of the abundant species. sively in remnants), 38 species in saltbush cover in remnants in both spring and Analyses for both plant and bird data sites (10 were exclusively in saltbush) and autumn and were absent from all other were conducted on a resemblance matrix 35 species in pasture sites (11 were exclu- treatments. of Bray–Curtis similarity measures (Clarke sively in pasture) across both seasons The analysis of the vegetation data using & Gorley 2006). Graphical analyses of the (Table 1). nMDS revealed that the vegetation assem- relationships between replicate samples Mean plant species richness (nati- blages were generally well separated within the three treatments and across the ve + exotic) ranged from 23.4 per quadrat across treatments with the pasture being two seasons were examined using the non- in remnant vegetation treatments in spring most dissimilar to the remnant vegetation metric multi-dimensional scaling (nMDS) 2008 to 3.4 per quadrat in saltbush treat- and the saltbush intermediate to the two routine. ments in autumn 2009 (Table 1). There (Fig. 2). There was also evidence of differ- To determine if there were differences were significant differences in species rich- ences in vegetation assemblages across sea- in plant and bird composition and abun- ness (ANOVA, P < 0.05) between treat- sons within each of the treatments and a dance resulting from treatment and ⁄ or sea- ments across seasons and between seasons close similarity between the saltbush and sonal effects, the data were analysed using across all treatments, although there was pasture treatments in the spring season the PERMANOVA+ procedure within PRI- no significant interaction between main (Fig. 2). MER v6. PERMANOVA+ is a permutational effects. The two-way crossed PERMANOVA+ multivariate analysis of variance that is free Mean plant species richness was signifi- analysis confirmed the pattern evident in from the assumptions of traditional para- cantly higher in remnants than in saltbush the nMDS. There were significant differ- metric ANOVA procedures and allows or pasture in both seasons. Mean species ences between treatments (F = 7.47, analysis of complex experimental designs richness in saltbush was significantly P = 0.0001) and seasons (F =10.61, (Anderson 2001; McArdle & Anderson higher than that in pasture in spring 2008, P = 0.0001), although the significant inter- 2001). The data were analysed as a 2-factor but not significantly different in autumn action term (F = 2.07, P = 0.0193) sug- fixed model design, with season and treat- 2009 (aprioricontrasts, P <0.05). gested that these differences are ments as factors. The Monte Carlo permu- Changes in the vegetation structure dependent on which treatment and season tation procedure within PERMANOVA+ (ground cover and plant life form cover) combination is examined. was used to calculate the significance in are evident from photo point images taken Subsequent pair-wise tests revealed that each analysis. Where differences were in spring 2008 and autumn 2009 (Fig. 1). within the autumn season all treatments identified, pairwise comparisons were con- Grazing of the saltbush sites in late were significantly different from one ducted for treatments within each season, summer and early autumn led to almost another. In the spring season, the rem- and seasons within each treatment. complete defoliation of the saltbush plants nants were different from both the salt- Where PERMANOVA+ revealed signifi- and the removal of most of the ‘volunteer’ bush and pasture, but the saltbush and cant differences between treatments or herbaceous ground cover components pasture treatments were not significantly seasons, the bird species most responsible between saltbush rows. different. The pair-wise tests across

ª 2011 Ecological Society of Australia ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 39 40 REPORT RESEARCH CLGCLMNGMN ETRTO O 2N PI 2011 APRIL 1 NO 12 VOL RESTORATION & MANAGEMENT ECOLOGICAL

Table 1. Summary of vegetation data recorded from five replicate sites (30 · 30 m) in remnant, saltbush and pasture treatments in spring 2008 and autumn 2009

Cumulative plant species richness Treatment Remnant Saltbush Pasture All treatments† Season Spring 08 Autumn 09 Spring 08 Autumn 09 Spring 08 Autumn 09 Spring 08 Autumn 09 Grand Total Native species 31 17 4 2 3 2 33 19 35 Exoticspecies192326 2311411547 Unknown71101071 7 All species 57 20 37 8 27 13 81 35 – Total species richness 58 [39] 38 [10] 35 [11] 89 (spring + autumn)‡ Mean species richness§ 23.4 (1.5)a 7.8 (0.6)b 17.0 (1.1)c 3.4 (0.4)d 10.0 (2.5)e 4.0 (1.3)d –– – Ground cover¶ Leaf litter 16.2 (3.7) 36.0 (6.0) 21.2 (10.8) 32.8 (14.6) 15.8 (6.3) 29.4 (13.0) – – – Microphytic crust 28.8 (6.7) 32.0 (5.1) 1.0 (0.0) 0.0 (0.0) 0.0 (0.0) 0.0 (0) – – – Native cover 2.2 (0.8) 2.4 (1.1) 1.0 (0.0) 0.2 (0.2) 1.0 (0.0) 4.0 (4.0) – – – Exotic cover 11.8 (4.7) 1.6 (0.9) 38.0 (8.6) 1.8 (0.8) 57.6 (14.6) 10.2 (8.7) – – – Plant life forms Tall mallee 13.8 (4.4) 18.6 (6.9) –– – Small mallee 10.5 (2.1) 10.5 (1.7) –– – Tall shrubs 17.8 (6.0) 27.2 (10.5) –– – Medium shrubs 3.5 (1.5) 1.6 (0.7) 34.0 (9.1) 34 (9.8) – – – Small shrubs 1.0 (0) 7.0 (0.0) 2.2 (2.0) – – – Herbs 8.3 (2.9) 1.8 (0.9) 17.0 (1.2) 1.8 (0.8) 30.2 (8.0) 8.2 (4.1) – – – Mat plants 1.0 (0) –– – Low grasses 6.8 (3.5) 1.6 (0.9) 17.4 (4.2) 0.8 (0.2) 38.0 (3.4) 6.4 (5.9) – – – Low tussocks 1.5 (0.5) 0.4 (0.2) 2.0 (1.0) – – – Vines, Climbers 1.0 (0) 0.8 (0.2) 1.0 (0) – – – Count of plant life forms 10 8 6 4 2 2 – – – †This column contains cumulative totals of species recorded across all treatments. ‡This row contains cumulative totals of species recorded in both spring 2008 and autumn 2009. Values

ª in square parentheses [ ] are the number of plant species recorded exclusively in each treatment. §Species richness is given as mean of five samples per treatment (±1 standard error). 01Eooia oit fAustralia of Society Ecological 2011 Values sharing the same letter are not significantly different ¶Ground cover components are given as mean percentage cover of five samples per treatment (±standard error). RESEARCH REPORT

were found only in saltbush sites and one bird species, the Brown Songlark (Cinclor- hamphus cruralis) was found only in pasture sites across both seasons. Nine spe- cies were recorded in at least one site across all three treatments (Appendix S2). Four species listed as ‘Rare’ under Sche- dule 9 of the South Australian National Parks and Wildlife Act 1972 were recorded during the survey period. The Elegant Parrot was recorded in two of the Figure 1. Photo point of the same saltbush planting prior to grazing (Spring 2008) and after saltbush sites in spring; the Hooded Robin grazing (Autumn 2009). The right hand photo taken in 2009 shows that an almost complete defolia- (Melanodryas cucullata cucullata)was tion of saltbush plants and a reduction in volunteer groundcover occured following grazing. recorded in one remnant site in spring and in one saltbush site in autumn; the Rest- less Flycatcher (Myiagra inquieta)was 2D Stress: 0.09 recorded in one remnant site in spring and autumn and one saltbush site in spring; and the White-winged Chough (Corcorax melanorhamphos) was recorded in only one remnant site in autumn (Appendix S2). Two introduced bird species, the Common Starling (Sturnus vulgaris) and the House Sparrow (Passer domesticus)were recorded infrequently across all treatments in both seasons. Despite the coarse seasonal differences evident from Table 2, there were no statis- tically significant differences between sea- Figure 2. Two-dimensional ordination (non-metric multi-dimensional scaling) of plant pres- sons (across all treatments) for bird species ence ⁄ absence data showing spring (shaded) and autumn (open) samples within each treatment. richness (ANOVA, F = 2.45, P > 0.05). m ⁄ D =remnant,d ⁄ s =saltbush, ⁄ h= pasture. However, using combined seasonal data, there were significant differences across all seasons within each treatment revealed sig- pasture treatments contained a total of 50, treatments for species richness (ANOVA, nificant seasonal differences (P <0.05)in 24 and 12 species respectively across both F = 26.04, P < 0.05). Mean species rich- vegetation from spring to autumn in all seasons. Five species that were recorded in ness was significantly higher in remnant treatments. the autumn surveys were not recorded in sites than in saltbush or pasture. Saltbush spring and 22 species that were recorded sites had significantly higher mean bird Bird community composition during spring surveys were not recorded species richness than pasture sites and structure during autumn (Table 2). (P < 0.05, Fig. 3). A total of 57 bird species was observed Thirty-two bird species were recorded There was a significant difference in during spring and autumn surveys (Appen- only in remnant sites, four bird species: overall bird abundance between seasons dix S2). Of these species, 52 were the Elegant Parrot (Neophema elegans), (ANOVA, F = 6.89, P < 0.05) and across all recorded in spring 2008 and 35 species Singing Honeyeater (Lichenostomus vires- treatments (ANOVA, F = 8.88, P <0.05). were recorded during autumn 2009 cens), Magpie-Lark (Grallina cyanoleuca) There was no significant difference in total surveys (Table 2). Remnant, saltbush and and Australian Kestrel (Falco cenchroides) bird abundance between remnant and salt- bush sites and remnant and pasture sites in spring 2009 (P > 0.05). In spring, pasture Number of bird species (native and introduced) recorded in all remnants, saltbush and Ta ble 2 . sites had significantly lower total bird abun- pasture sites in spring 2008 and autumn 2009. Numbers in parentheses indicate the number of spe- dance than saltbush sites (P < 0.05), with cies found exclusively in each treatment. Row and column totals are cumulative similar trends apparent across treatments Remnant Saltbush Pasture Total (cumulative) in autumn 2009. Spring 2008 45 (33) 17 (4) 10 (1) 52 (22) The analysis of the avian data using Autumn 2009 29 (19) 15 (3) 6 (1) 35 (5) nMDS revealed that the bird assemblages Total (spring and autumn) 50 (32) 24 (4) 12 (1) 57 were generally separated across land

ª 2011 Ecological Society of Australia ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 41 RESEARCH REPORT

16 resources for a diverse range of native fauna. For example, biological surveys of a 12 the Olary Plains in South Australia found mammal, bird and reptile groups that are 8 b associated with different types of cheno- pod shrubland (Forward & Robinson 4 c 1996; Playfair et al. 1997). The depen- Number of species dence of some native species on shrubby 0 Remnant Saltbush Pasture vegetation systems suggests that planting such systems in agricultural areas may at Figure 3. Mean (±1 standard error) values for bird species richness (combined seasonal data) least partially satisfy the habitat and in remnant, saltbush and pasture treatments. Means sharing the same letter are not significantly resource requirements of these species. different (aprioricontrasts P >0.05). Changes to less intensive land manage- ment, coupled with predicted climate vari- management treatments. There was how- A similar suite of bird species contrib- ability, may also provide opportunities for ever, no strong indication of differences uted to a large proportion of the dis- new species to colonise or move through across seasons within each treatment similarity between sites in the different the landscape. (Fig. 4). treatments and seasons (Appendix S3). Despite the fact that the study area PERMANOVA+ confirmed the general Species contributing most to the dissimilar- lies approximately 100 km south of the pattern displayed in the nMDS, revealing ity between two treatments were typically natural distribution of Old Man Saltbush, significant differences between bird assem- present in both treatments, with differ- a number of native bird species were blages in the different land use treatments ences due to higher or lower average abun- recorded in saltbush plantings that were (F =6.47,P = 0.0001) but no seasonal dif- dance. Exceptions to this pattern were not observed in adjacent areas of pasture. ferences (F =0.77, P = 0.18). There was those species restricted to remnant sites It is uncertain whether these species are however, a significant treatment by season such as the Yellow Thornbill (Acanthiza resident in the saltbush patches or if they interaction term (F =1.82, P = 0.0185), nana), Common Bronzewing (Phaps chal- are opportunistically and ⁄ or temporarily indicating that differences in treatments coptera), Spotted Pardalote (Pardalotus using resources provided by the saltbush. were dependent on the season in which punctatus) and Red Wattlebird (Anthoch- The current findings suggest that ele- the surveys were conducted. aera carunculata). ments of these planted systems may pro- Pairwise tests of the levels of treatment vide at least partial habitat for native bird within season revealed that the bird assem- species, compared with more structurally blages in the saltbush and pasture treat- Discussion simple and ⁄ or complex landscape ele- ments were not significantly different in Potential biodiversity values ments. autumn (P > 0.05). Bird assemblages in all of saltbush Of particular interest was the observa- other treatments were significantly differ- tion of nesting Orange and White-fronted ent in both the spring and autumn seasons Naturally occurring chenopod shrublands Chats in only saltbush sites during the (P <0.05). in semi-arid areas provide habitat and spring 2008 survey, providing evidence of the potential value of these areas for native 2D Stress: 0.16 bird species that are naturally associated with shrub-layer vegetation. Seddon et al. (2009) also observed two species of chat (Orange and Crimson) feeding and shelter- ing in planted saltbush alleys in non-saline areas of the NSW Central Western Plains. The high abundance of such eruptive bird species is likely to change seasonally and annually, depending on climatic conditions in the study landscape and in surrounding areas. Conditions during sampling were typical of the seasonal climatic patterns of the region. Figure 4. Two-dimensional ordination (non-metric multi-dimensional scaling) of fourth root Most of the bird species recorded in transformed bird abundance data showing each treatment. m ⁄ D =remnant, d ⁄ s = saltbush, saltbush were also recorded in one or n ⁄ h = pasture. Shaded shapes are spring 2008 surveys and unshaded shapes are autumn 2009 both of the other two land uses, suggest- surveys. ing a degree of plasticity in their

42 ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 ª 2011 Ecological Society of Australia RESEARCH REPORT behaviour and resource requirements in be the reason for this difference (Collard Management considerations relation to the agricultural matrix. A large et al. 2009a). proportion of these species (42%) were Saltbush plantings are primarily established Resources available to birds habitat ‘Generalists’ (sensu Attwood et al. by land managers on non-saline lands in in planted saltbush at 2009) with very few ‘Woodland special- southern Australia as fodder crops. Intense different spatial and ists’ observed in saltbush sites. The habitat grazing in these systems, typically during temporal scales requirements of the latter species are unli- late summer and early autumn, is reflected kely to be met by a grazed, monoculture The preliminary observations from this in the large seasonal differences shown by of saltbush. Similarly, the small, isolated study suggest that planted saltbush vegeta- plant and bird communities in the present and modified remnants sampled in this tion may play a role in providing resources study. Some of these differences may be study contained fewer species than could for a range of native bird species, some of due to the effects of season alone, rather be expected in larger and less disturbed which are of conservation significance. than grazing management or a combina- vegetation remnants in the landscape (e.g. Collard and Fisher (2010) propose some tion of both. Saltbush plants are able to Billiat or Ngarkat Conservation Parks). potential mechanisms by which planted recover from this annual heavy grazing The observations of three threatened fodder shrub systems can provide benefits pressure, however the prolonged reduc- bird species (the Elegant Parrot, Hooded to native fauna at different spatial scales. tion in vegetative structure and ground- Robin and Restless Flycatcher) and a range These benefits depend on the sheltering, cover may affect other plant and animal of other native birds in saltbush plantings foraging and nesting requirements of indi- species in different ways. For example, in this study support the notion that these vidual species at a site scale and the mobil- bird species associated with dense, ‘novel’ ecosystems (sensu Hobbs et al. ity, phenotypic plasticity and resilience of shrubby vegetation (e.g. Inland Thornbill, 2008) may also complement existing frag- populations at a landscape scale. On-going Acanthiza apicalis) may be displaced or mented vegetation communities and thus investigations are underway on the specific unable to survive while those favouring potentially contribute to regional biodiver- resources that planted saltbush offers to more open structure (e.g. Australian Mag- sity conservation. These threatened spe- different bird species and whether the pie, Cracticus tibicen) may benefit. The cies are not generally associated with plantings provide more than just an oppor- timing and intensity of grazing in saltbush shrubland habitat, rather they are typically tunistic resting stop or stepping stone could be managed to better suit the found in habitat containing trees. Their between preferred habitats. Furthermore, requirements of some of these species, presence in saltbush sites and complete seasonal differences in bird community particularly those of higher conservation absence from adjacent pasture sites, sug- composition suggest temporal variability significance. gests that the saltbush may be providing and it is unclear how individual species We have shown that saltbush plantings them resources in the form of food, shelter respond to this. can enhance vegetation structural com- or structure for perching. These threa- Site-level attributes and wider land- plexity compared with existing pasture tened bird species exhibit features and scape context are both important deter- systems and thus potentially provide behaviours consistent with ‘Woodland minants of species’ occurrence in resources for native fauna. Furthermore, generalists’ proposed by Attwood et al. agricultural landscapes (Lindenmayer & ‘volunteer’ or planted groundcover com- (2009). Hobbs 2004; Collard et al. 2009b). In par- ponents between saltbush rows have the Our findings are similar to those of Sed- ticular, structural complexity is an impor- potential to improve stock-carrying capac- don et al. (2009) who showed higher tant factor affecting the occurrence and ity, reduce soil erosion and enhance floris- species richness of birds in remnant vege- abundance of different fauna species (Fi- tic diversity (Norman et al. 2008). The tation compared with three year old salt- scher et al. 2004; Munro et al. 2007). majority of the volunteer species present bush alley plantings in central western Findings from the present study suggest in the sampled saltbush plantings, includ- New South Wales. Unlike these authors that the higher structural complexity of ing Old Man Saltbush, was exotic in origin who found no difference in the number of saltbush plantings (when compared with and a number of them were declared bird species between saltbush and crop surrounding intensively managed pas- weeds in South Australia. Whilst they may rotation (conventional) sites, we found sig- tures) and changes to vegetation structure provide some resources for native fauna, nificantly higher bird species richness in caused by seasonal and ⁄ or management consideration needs to be given to the pos- saltbush than in pasture treatments. Sed- influences may have a significant effect on sible consequences of weeds and poten- don et al. (2009) also sampled saltbush bird communities. As identified by other tially other vertebrate and invertebrate alleys (i.e. 15 m wide saltbush strips, alter- authors (e.g. Fischer et al. 2004; Munro pests in saltbush plantings and the impacts nating with 50 m wide strips of pasture), et al. 2007), more information is needed of pest management on biodiversity values whereas contiguous blocks greater than on landscape factors as well as the habitat at local and regional scales. 2 ha were sampled in the present study. and resource requirements of different Environmental benefits of saltbush The floristic diversity and structural com- species to better inform the placement of plantings may be enhanced by other man- plexity of the saltbush plantings compared planted woody perennial systems in farm- agement improvements such as changes to with adjacent areas of pasture is likely to ing landscapes. the timing and intensity of grazing or by

ª 2011 Ecological Society of Australia ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 43 RESEARCH REPORT incorporating more than one fodder shrub sampling sites; Rowena Danks (Murray-Mal- Fischer J., Lindenmayer D. B. and Fazey I. (2004) species. For example, mixed shrub-based lee Local Action Planning Group), Ray Appreciating ecological complexity: habitat contours as a conceptual landscape model. forage systems are currently being exp- O’Malley, Greg Sweeney (SARDI) and Mike Conservation Biology 18, 1245–1253. lored as viable enterprises that are poten- Bennell for assistance with site selection; Forward L. R. and Robinson A. C. (1996) A Biolog- tially more resilient in the face of climate Department for Environment and Heritage ical Survey of the South Olary Plains, South Australia, 1991–1992. Biological Survey and change (Hobbs et al. 2009). The high pro- for supplying aerial photography; Ann- Research, Natural Resources Group, Depart- ductivity of saltbush on low fertility soils Marie Mabarrack and Kate Willing (Rural ment of Environment and Natural Resources, also makes it a potentially useful species Solutions SA) for GIS and mapping sup- Adelaide, SA. Hobbs R. J. (1993) Effects of landscape fragmen- for biosequestration in the light of emerg- port; Zita Stokes (Rural Solutions SA) for tation on ecosystem processes in the western ing carbon markets (Hobbs 2009). her useful comments. Thanks to staff from Australian wheatbelt. Biological Conservation the Land and Biodiversity Conservation 64, 193–201. Conclusion and future Hobbs T. J. (2009) Regional industry potential for Division (within the former DWLBC) for directions woody biomass crops in lower rainfall south- their support and technical assistance. ern Australia. FloraSearch 3c. Report to the The results presented here contribute to a Funding for a significant amount of this Joint Venture Agroforestry Program and Future Farm Industries CRC. Publication No. greater understanding of the value of work was provided through the SA 09 ⁄ 045. Rural Industry Research and Devel- planted saltbush systems for biodiversity, Complementary State NRM Program opment Corporation, Canberra, ACT. compared with other better known land- 2008 ⁄ 2009 and the Future Farm Industries Hobbs R. J., Arico S., Aronson J. et al. (2008) Novel ecosystems: theoretical and manage- scape elements. We sampled only five sites Co-operative Research Centre. ment aspects of the new ecological world in each land management treatment over order. Global Ecology and Biogeography 15, the course of a year within a relatively small 1–7. References Hobbs T. J., Bennell M. and Bartle J. (2009) area of the South Australian Murray Mallee. Developing species for woody biomass crops There was also considerable variation in Anderson M. J. (2001) A new method for non- in lower rainfall southern Australia. Flora- parametric multivariate analysis of variance. grazing management of saltbush and pas- Search 3a. Report to the Joint Venture Agro- Austral Ecology 26, 32–46. forestry Program and Future Farm Industries ture sites, as well as differences in time Attwood S. J., Park S. E., Maron M. et al. (2009) CRC. Publication. No. 09 ⁄ 043. Rural Industry since establishment for saltbush. Despite Declining birds in Australian agricultural Research and Development Corporation, landscapes may benefit from aspects of the these experimental inconsistencies, statisti- Canberra, ACT. European agri-environment model. Biological Lefroy E. C. and Smith P. F.(2004) The biodiversity cally significant differences between treat- Conservation 142, 1981–1991. value of farming systems and agricultural land- ments and seasons were detected for both Barrett G., Silcocks A., Barry S., Cunningham R. scapes. Pacific Conservation Biology 10, 80– and Poulter R. (2003) The New Atlas of Aus- plant and bird communities. 87. tralian Birds. Royal Australasian Ornitholo- Lefroy E. C., Flugge F., Avery A. and Hume I. Opportunities exist to further quantify gists Union, Melbourne. (2005) Potential of current perennial plant- the biodiversity values of saltbush plant- Bartel B. and Knight A. (2000) Old Man Saltbush: based farming systems to deliver salinity man- Farmer Experiences in Low Rainfall Systems. ings in the study landscape and elsewhere. agement outcomes and improve prospects for Primary Industries and Resources South Aus- native biodiversity: a review. Australian Jour- Subsequent work in the Murray Mallee is tralia, Adelaide. nal of Experimental Agriculture 45, 1357– beginning to uncover details about how Clarke K. R. and Gorley R. N. (2006) PRIMER v6: 1367. User Manual ⁄ Tutorial. PRIMER-E, Plymouth. birds are actually using the resources on Lindenmayer D. B. and Hobbs R. J. (2004) Fauna Collard S. J. and Fisher A. M. (2010) Shrub-based conservation in Australian plantation forests – offer in these saltbush systems (e.g. food, plantings of woody perennial vegetation in a review. Biological Conservation 119, 151– shelter, feeding substrates, nesting mate- temperate Australian agricultural landscapes: 168. what benefits for native biodiversity? Ecologi- rial). Combining data on local-scale attri- McArdle B. H. and Anderson M. J. (2001) Fitting cal Management and Restoration 11, 31–35. multivariate models to community data: a butes (e.g. groundcover, plant life forms, Collard S. J., Fisher A. M. and McKenna D. J. comment on distance-based redundancy grazing intensity) with landscape context (2009a) Biodiversity value of saltbush (Atri- analysis. Ecology 82, 290–297. plex nummularia) plantings in mixed farming information (e.g. composition and configu- McKenna D., Bennell M. and Mazanec R. (2009) landscapes of the Southern Mallee, South Review of Atriplex nummularia Old Man Salt- ration of surrounding landscape elements), Australia. DWLBC Technical Report 2009 ⁄ 27, bush and its potential for domestication as a may also help to explain the observed pat- Government of South Australia, through fodder plant in Australian dryland farming sys- Department of Water, Land and Biodiversity terns in faunal communities in planted salt- tems. In: Reviews of High Priority Species for Conservation, Adelaide. Woody Biomass Crops in Lower Rainfall bush systems. Such information would be Collard S. J., LeBrocque A. F. and Zammit C. Southern Australia – FloraSearch 3b (eds T. J. useful to guide future decisions made by (2009b) Bird assemblages in fragmented agri- Hobbs, J. Bartle and M. R. Bennell), pp. 97– cultural landscapes: the role of small brigalow land owners and regional natural resource 139. RIRDC Publication No. 09 ⁄ 044. Rural remnants and adjoining land uses. Biodiversity Industries Research and Development Corpo- management bodies on the strategic place- and Conservation 18, 1649–1670. ration, Barton, ACT. ment and on-ground management of Croft S. J., Pedler J. A. and Milne T. I. (2005) Millsom D. A. (2002) Direct seeding of saltbush: Bushland Condition Monitoring Manual: perennial fodder shrubs. landholder-driven initiatives. Ecological Man- Southern Mount Lofty Ranges. The Nature agement and Restoration 3, 156–166. Conservation Society of South Australia, MMLAP (2001). Murray Mallee Revegetation Adelaide, SA. Plan. Murray Mallee Local Action Planning Acknowledgements DWLBC (2008) Re-establishment of vegetation in Association Inc., Murray Bridge. South Australia 2008. DWLBC Report Morrison M., Durante J., Greig J. and Ward J. We are grateful to the following peo- 2009 ⁄ 11. Department of Water, Land and Bio- (2008). Encouraging Participation in Market ple: landholders who allowed access to diversity Conservation, Adelaide.

44 ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 ª 2011 Ecological Society of Australia RESEARCH REPORT

Based Instruments and Incentive Programs, (eds D. Lindenmayer, S. Dovers, M. Harriss Appendix S2. Bird species recorded Land and Water Australia, Canberra, ACT. Olson and S. Morton), pp. 19–25. CSIRO Pub- across all sampled land uses and seasons. Munro N. T., Lindenmayer D. B. and Fischer J. lishing, Collingwood, Vic. (2007) Fauna response to revegetation in agri- Prober S. M. and Smith F.P. (2009) Enhancing bio- Appendix S3. Average dissimilarity bet- cultural areas of Australia: a review. Ecologi- diversity persistence in intensively used agri- ween treatments and the average abun- cal Management and Restoration 8, 199–207. cultural landscapes: a synthesis of 30 years of dance of the bird species contributing up Newton P. J. and Yunusa I. A. (2002) Future roles research in the Western Australian wheatbelt. for native woody species in Australian agricul- Agriculture, Ecosystems and Environment to 30% of the dissimilarity between treat- tural landscapes. In: Agriculture for the Aus- 132, 173–191. ments. Species are listed in decreasing tralian Environment. Proceedings of the 2002 Seddon J., Doyle S., Bourne M., MacCallum R. order of their importance in discriminating Fenner Conference on the Environment (eds and Briggs S. (2009) Biodiversity benefits of B. P. Wilson and A. Curtis), pp. 326–338. alley farming with old man saltbush in central between the two sets of samples. Johnstone Centre, Charles Sturt University, western New South Wales. Animal Production Please note: Wiley-Blackwell are not res- Albury, NSW. Science 49, 860–868. ponsible for the content or functionality of Norman H. C., Masters D. G., Silberstein R. et al. Smith F. P. (2009) Assessing the habitat quality of (2008) Sustainable Grazing on Saline Lands oil mallees and other planted farmland vegeta- any supporting information supplied by WA 1 research project – final report. Available tion with reference to natural woodland. Eco- the authors. Any queries (other than miss- from URL: http://www.landwaterwool.gov. logical Management and Restoration 10, ing material) should be directed to the au/products/pk071404. 217–227. Playfair R. M., De Jong M. C. and Matthew J. S. corresponding author for the article. (1997) North Olary Plains Biological Survey: birds. In: A Biological Survey of the North Supporting Information Olary Plains South Australia, 1995–1997 (eds R. M. Playfair and A. C. Robinson), pp. 139– Additional Supporting Information may be 150. Natural Resources Group, Department of found in the online version of this article. Environment and Natural Resources, Ade- laide, SA. Appendix S1. Plant species recorded in Prober S. M. and Hobbs R. J. (2008) Temperate each land use treatment. eucalypt woodlands. In: Ten Commitments: Reshaping the Lucky Country’s Environment

ª 2011 Ecological Society of Australia ECOLOGICAL MANAGEMENT & RESTORATION VOL 12 NO 1 APRIL 2011 45