Management of remnant lowland grasslands and grassy woodlands for nature conservation: a review Ian D. *

Abstract Collectively, they encompassed about a A review is presented of recent third of the State (Fig. I). Sadly, the des- literature relevant to the conservation truction of grassy ecosystems has been so management of remnant lowland complete that today less than 0.5 percent grasslands and grassy woodlands, two of remain (Ministry for Planning and the most threatened ecosystems in temp.. Environment 1989). The situation is erate Australia. The long-term conserva- similar in New South Wales (Sim and tion of both ecosystems is dependent upon Urwin 1984),the Australian Capital Terri- increased reservation and informed tory (Groves 1979), South Australia management. Issues discussed include (Davies 1982, Williams and Goodwins grazing by stock and native herbivores, 1988) and Tasmania (Kirkpatrick et al. burning, ploughing, grading, mowing, 1988). The decimation of grassy eco- trampling, tree regeneration and planting, systems is not a peculiarly Australian herbicides, fertilizers, rehabilitation and phenomenon, but echoes the loss to agri- ecosystem re-creation. culture of the indigenous grasslands of Introduction other continents, including the prairies of Lowland (non-alpine) grasslands and North America, the pampas of South grassy woodlands are among the most America and the chalk grasslands of threatened ecosystems in Australia, with Europe (Koopowitz and Kaye 1983,Davis few examples in conservation reserves et al. 1986, Murphy and Ehrlich 1989). (Specht 1981). These communities Mack,(1989)observed that: originally covered vast areas of the low- 'In less than 300 years (and in most land plains of temperate, south-eastern cases, little more than 100years) much Australia. However, the fertile plains were of the temperate grassland outside the first areas to attract European settle- Eurasia ... has been irreparably ment, and agriculture and associated transformed by human settlement and development rapidly transformed the the concomitant introduction of alien native ecosystems to pastures, crops and plants. Few other changes in the towns. distribution of the earth's biota since Throughout south-eastern Australia, the end of the Pleistocene have been as native grasslands and grassy woodlands radical ... or as swift'. are now restricted to small, fragmented Unless remedial action is undertaken remnants, the best examples of which promptly, many of the distinctive com- commonly occur on rail and road ease- munities of lowland grasslands and grassy ments (Groves 1979, Scarlett and Parsons woodlands in , and many of the 1982, Frood and Calder 1987). In rare and threatened species they contain, Victoria, native grasslands and grassy will soon become extinct. A compre- woodlands originally dominated the hensive conservation program must Wimmera and Northern Plains, Dundas immediately be implemented if these Tablelands, Western Volcanic Plains, threatened ecosystems and species are to Gippsland Plains and other, smaller areas. survive. To succeed, such a program must include (1) the protection and mainten- ·Flora Group, Department of Conservation and ance of intact remnants, including those Environment. P.O. Box 406. Kew,3101 Present address: Botany Department, laTrobe containing rare or threatened species, (2) University. Bundoora, 3083 the establishment, in all regions of the

56 Victorian Nat. Fia. 1. Probable distribution of lowlandasslands~ and grassywoodlands prior to European settlement of Victoria.The mapped area In the Raak Plain in north-westVictoriacomprisesa mosaic of grasslandsand saline shrublands. Additional areas may occur in south Gippsland. Based on land-systemmaps prepared by Rowan (1988)for the Land Conservation Council (1988),seetext for details. state, of reserves large enough to remain management; the latter topic deserves a viable, and (3) the restoration of plant review of its own. The paper is an ex- communities and the re-estabiishnient of tended version of one presented earlier rare plants and animals within these (Lunt 1990a). Plant names follow Ross secure reserves (Frood and Calder 1987, (1990). Lunt in press). Unlike many vegetation types, lowland Although a comprehensive network of grasslands and grassy woodlands often re- large reserves is vital, reservation is but quire intensive management to maintain the first step in conserving these their biological values. Most Themeda threatened ecosystems. Ultimately, their grasslands, for instance, require burning conservation depends upon appropriate (or some other form of canopy removal) management. This problem is graphically every three to five years to prevent the illustrated in Hattah-Kulkyne National vigorous Kangaroo Grass (Themeda , where intensive· grazing by large triandra) from smothering smaller native populations of Western Grey Kangaroos herbs (Stuwe and Parsons 1977, (Macropus fuliginosus) continues to McDougaIl1989a). degrade 'protected' plant communities such as Pine-Buloke woodland, by Ecosystem definition consuming and preventing the Lowland grasslands and grassy wood- regeneration of many species, including a lands are non-alpine ecosystems, in which number of rare plants (Cheal 1986,Mueck mature trees are absent or scattered 1988). (generally with less than 30% cover). The The aim of this paper is to review recent ground layer is dominated by native literature relevant to the conservation grasses, and herbaceous plants management of lowland grasslands and predominate. Few ferns occur, and small- grassy woodlands. The paper is not in- leaved 'heathy' shrubs are typically un- tended to provide pithy prescriptions for common. 'Typical' Australian families site management. Attention is focussed on such as Myrtaceae, ProteaCeae and Epa- flora and vegetation rather than fauna cridaceae are poorly represented. Lowland

Vol. 108 No.3 (1991) 57 grasslands and grassy woodlands are most grassy woodlands. Within four years of extensive in flat to gently undulating European settlement, Robinson (in Bride landforms, on relatively fertile, heavily- 1898, Conley 1984) reported that Kang- textured soils. Collectively, 'lowland aroo Grass was replaced by a 'silk-grass' grasslands and grassy woodlands' (thought by Moore (1959) to be the exotic correspond to the 'Savannah' land sys- Squirrel-tailed Fescue, Vulpia bromoides). tems, which Specht (1972) described trom The ecological effects of grazing in- South Australia. In Victoria, these crease exponentially with grazing intensity ecosystems are almost totally confined to (Robinson and Dowling 1976). Morcom the lowland plains below 500 m altitude, (1990) demonstrated a progressive except for a series of isolated occurrences reduction in the number of native plant in inter-montane basins in the east of the species and a concomitant increase in the State, as at Buchan, Benambra and Omeo. cover and percentage of exotic species, with increasing grazing intensity. Original distribution As grazing pressure increases, tall, It is difficult to accurately map the pre- native perennial grasses such as Kangaroo European distribution of lowland grass- Grass and Silky Blue-grass (Dichantheum lands and grassy woodlands due to the sericeum) are replaced by small tussock- paucity of historical records and the near grasses such as Spear-grasses (Stipa total destruction of the ecosystems. The species) and Wallaby-grasses (Danthonia lack of recent vegetation surveys com- species), and then by introduced, annual pounds the problem. Recent maps of grasses and herbs (Moore 1962, 1964; Victorian vegetation (e.g. Paine 1982, Robinson and Dowling 1976).The overall Forests Commission Victoria 1984, Land change is from tall, summer.growing, Conservation Council 1988) either fail to perennial, native grasses to short, winter- distinguish lowland grasslands from growing, annual, introduced species. Such cleared pastures or fail to distinguish changes are accentuated by soil grassy woodlands from sclerophyllous for- disturbance and the application of fertil- ests, particularly in central Victoria. The izers (Robinson and Dowling 1976). map of lowland grasslands and grassy Grazing may also cause genotypic changes woodlands shown in Fig. 1is derived from within grazing-tolerant species, for ex- land-system maps prepared by Rowan ample, by selecting for short over tall (1988) for the Land Conservation Council Wallaby-grasses(Scott and Whalley 1984). (1988). It shows those areas which are Grazing by sheep is mote destructive than most likely to have supported grassy eco- grazing by cattle (Moore and Biddiscombe systems prior to settlement: principally, 1964) as sheep are more selective of pala- Rowan's (1988) 'plains' land-forms of fine- table plant species and graze more textured soils. Relevant vegetation maps uniformly and closely than cattle (Wells (e.g. Land Conservation Council 1987) 1969). and regional land system reports (e.g. Although Kangaroo Grass is sensitive Gibbons and Downes 1964) were also to grazing, and may rapidly decline under consulted. As shown in Fig. 1, lowland grazing pressure, it is not particularly pala- grasslands and grassy woodlands occupied table to stock except when it is young about eight million hectares, over a third (Leigh and Holgate 1978, Whalley etal. of Victoria, prior to European settlement. 1978). Thus, initial grazing pressure is often directed, not at the dominant grasses Grazing by stock such as Kangaroo Grass, but at less Grazing by domestic stock has had an common components of the flora, particu- incalculable impact on the composition larly forbs (Leigh and Holgate 1978). and structure of lowland grasslands and Species such as Yam Daisy (Microseris

58 Victorian Nat. lanceolata) are particularly palatable and unaltered. Deferred and rotational grazing sensitive to grazing (Farrington and regimes, in which stock are allowed to Mitchell 1966). Thus, stock grazing even graze in sensitive areas from late summer at low intensities can have a major impact to mid-winter only (when few native on the grassland flora, with little obvious plants are flowering or rapidly growing), effect on the dominant grasses. This may enhance the conservation of lowland process is well documented in alpine grasslands and grassy woodlands in some grasslands. Alpine Tussock-grasses (Poa cases. However, grazing by stock is clearly species) are oflow nutritional quality, and, inappropriate in conservation reserves to maintain an adequate'diet, cattle rely and in intact grassy ecosystems on public upon less abundant, but more nutritious land as, at best, it may maintain the herbs, including the composites, Snow quality of a lowland grassland or grassy Daisy (Celmisia asteliifolia), Billy- woodland, but it can rarely, if ever, Buttons (Craspedia sp.) and Alpine improve it. The grazing of high quality Podolepis (Podolepis robusta) (van Rees remnants that previously have escaped 1982, van Rees and Holmes 1986). grazing will rapidly, and probably Grassland remnants on grazed private irreversibly, lead to a deterioration in properties on the western plains typically vegetation quality. have fewer native species than ungrazed (and regularly burnt) sites on railway Grazing by native herbivores reserves (Stuwe and Parsons 1977). Stock Prior to European settlement, native rarely' graze a site uniformly, and areas herbivores, including kangaroos and such as drainage lines tend to be grazed smaller macropods, undoubtedly played most intensively.Grazing pressure in these an important role in the ecology of low- areas is usually considerably higher than land grasslands and grassy woodlands. in the site as a whole. Pugging of the moist Grazing, browsing, trampling and digging soil causes further damage. all influenced plant survival and repro- Nevertheless, Themeda grasslands in duction. The mammalian fauna of particular require periodic disturbance of lowland grasslands and grassy woodlands the vegetation (not the soil) to prevent the has been grossly depleted since European vigorous Kangaroo Grass from settlement, and many of the smaller mam- smothering smaller plants (Stuwe and mals are now extinct (Bennett 1982, Parsons 1977). Many remnants that have Seebeck 1984).Nevertheless, native mam- not recently been grazed, burnt or other- mals still play an important role in many wise 'disturbed' possess dense Kangaroo grassy woodlands. Grass with few other native species Robertson (1985) studied the impact of (McDougall 1987, 1989a). Eastern Grey Kangaroos (Macropus Management techniques such as giganteus) in a grassy woodland with a burning are not always suitable or possible Themeda understorey at Gellibrand Hill to implement, particularly on agricultural near Melbourne. Grey Kangaroos were land, and, in some instances, grazing by found to be highly selective grazers, stock may be preferable to no manage- feeding almost exclusively on monocoty- ment at all (Scarlett and Parsons 1982, ledons, particularly grasses and the exotic Fensham and Kirkpatrick 1989). The irid, Common Onion-grass (Romulea conservation value of lowland grasslands rosea). Forbs and woody plants were not and grassy woodlands that form significantly grazed, despite their higher unimproved pastures on private proper- nutritive value. Such grazing had a greater ties are unl,ikelyto degrade rapidly if the selective impact on the, vegetation than current management regime (including did burning, and promoted herbs that stocking rates and fertilizer usage) remains kangaroos found unpalatable, such as the

Vol. 108 No.3 (1991) 59 exotic Cape Weed (Arctotheca calendula) species), which require fires in autumn, and clovers (Trifolium species). Grazing rather than late spring or early summer, in by Eastern Grey Kangaroos in the order to permit flowering and seed pro- TidbinbilIa Nature Reserve (ACT) was duction (Scarlett and Parsons 1982). found to deplete a number of herbs in- Regular burning in Themeda grasslands cluding Kangaroo Grass, and to favour maintains the diversity of native plants by species such as the native Red-leg Grass preventing the vigorous Kangaroo Grass (Bothriochloa macra) and the exotics, Soft from outcompeting smaller native plants Brome (Bromus hordeaceus ssp. mol/i- (Stuwe and Parsons 1977, Robertson formis), Cat's-ear (Hypochoeris radicata) 1985). However, burning also promotes and clovers (Trifolium species) (Neave and many exotic species, particularly in de- Tanton 1989). graded remnants (Lunt 1990b). By contrast, Swamp Wallabies (Wal- McDougall (1989a) provides a compre- labia bicolor) browse rather than graze, hensive review of recent literature on and generally eat woody and broad-leaved burning in Themeda grasslands, and the plants, plus a proportion of grasses. The reader is referred to that document for dominant food plants of Swamp Wallabies further discussion. McDougall's (1989a) at Gellibrand HilI are eucalypt seedlings, mor~ conclusions were: Hedge Wattle (Acacia paradoxa) and the 'Fire is necessary to prevent build-up exotic forb, Ribwort (Plantago lanceolata) Qf Themeda litter and maintain species (Allen 1987). diversity in natural and restored The impact of kangaroo grazing de- grasslands.' pends upon the population density of 'Burning should be done at least every kangaroos. Neave and Tanton (1989) con- five years but will probably not be cluded that kangaroo populations in detrimental if done as frequently as Tidbinbilla Nature Reserve should be every two years. The amount of fuel diminished in order to maintain a produced by a grassland in the first Themeda-dominated understorey. More year after burning will generally not be dramatically, large populations of Western enough to warrant control burning for Grey Kangaroos (Macropus fuliginosus) fire protection.' pose a serious threat to Pine-Buloke wood- 'The time of burning is probably not lands in Hattah-Kulkyne National Park. critical unless burning is done annually. As mentioned earlier, high population It will most often be regulated by ease levels severely degrade vegetation and of burning. Late summer and autumn restrict the regeneration of many rare and fires will therefore be most common.' threatened plants (Cheal 1986, 1988, A primary aim of management at many . Coulson and Norbury 1988, Mueck 1988). sites will be to protect rare or threatened plants. In such instances, burning should Burning only be conducted after such plants have Lowland grasslands and grassy wood- flowered and shed seed. Since most native lands were frequently burnt by Aborigines herbs in lowland grasslands and grassy before European settlement (Nicholson woodlands flower in mid-spring (Groves 1981). Fire still plays an important 1965) this may necessitate burning in late ecological role in many grassland summer or autumn. remnants, particularly those on railway It must be emphasised, however, that reserves. A number of endangered plants the conclusions above have been made on of lowland grasslands and grassy wood- the basis of the limited data available, and lands depend on specific burning regimes that further research on fire regimes in for regeneration, particularly late- lowland grasslands and grassy woodlands flowering native legumes (e.g. Psoralea is urgently required. Furthermore, should

60 Victorian Nat. these conclusions prove correct, they may in Themeda cover, but concluded that apply only to Themeda grasslands; intact, although somewhat degraded, species-rich Red Gum (Eucalyptus camal- Themeda grasslands were 'reasonably dulensis) and Yellow Box (E. melliodora) tolerant' of mowing. As a general rule, woodlands in the Grampians (Lunt mowing is probably less favourable to I99Oc), and Sandplain Grassland in the native grasslands than is burning Mallee, for instance, do not appear to need (Kirkpatrick et al. 1988).However, annual fire to maintain their diversity. or biennial mowing of Themeda grasslands may be preferable to no management at all. Chan (1980) recom- Ploughing and lP'ading mended that native grasslands not be Ploughing, grading, ripping, and other mown between early October and late soil disturbances are· antithetical to the January, to promote the growth, carbo- conservation of lowland grasslands and hydrate accumulation and floweringof the grassy woodlands. Hobbs and Atkins dominant grasses. Grassy ecosystems (1988) demonstrated that soil disturbance should not be mown to less than 10 em greatly favours the establishment of intro- high, in order to permit the survival of duced species. As noted by Stuwe (1986), native herbs (Lodder et al. 1986). 'ploughing or grading cause rapid, often virtually complete, removal of native vegetation and its replacement by Trampling and vehicular impact introduced species'. The small herbs of many lowland grass- lands and grassy woodlands are often Mowing susceptible to trampling, both by foot and Little research has been undertaken on vehicular traffic. Damage is most marked the effects of mowing or slashing in low- when vehicles drive over soft, moist soil, leaving indelible wheel ruts. Kangaroo land grasslands and grassy woodlands. Grass is very susceptible to trampling, Stuwe (1986) believed that 'slashing or particularly when old and un burnt mowing adds to plant litter and favours (McDougall I989a), and can be eliminated the ingress of introduced species via dis- from wheel ruts after a single drive over turbances caused by the machinery used soft, wet ground (Lunt 1987). and aggregated litter'. Similarly, (equivocal data in) Kirkpatrick (1986) suggested that mowing may diminish the diversity of Tree regeneration native plants. Regular mowing of grassy The density of trees in many areas of woodlands prevents tree regeneration. Victoria has declined considerably since Chan (1980) found that Kangaroo European settlement (Woodgate and Grass, Wallaby-grasses (Danthonia car- Black 1988) and many of the original phoides and D. auriculata) and Kneed woodlands are now treeless or only Spear-grass (Stipa bigeniculata) declined sparsely treed. Woodlands of Silver when mown every three months or more Banksia (Banksia marginata) were cleared often. Such frequencies also promoted a so soon after settlement that they were number of exotics, including Squirrel-tail omitted from many of the earliest Fescue (Vulpia bromoides), Hair-grasses surveyors maps (Powell 1967). Prior to (Aira species), Rib-wort (Plantago European settlement, tree densities were lanceolata), Clovers (Trifolium species), controlled by a range of factors, including Garden Dandelions (Taraxacum sect. climate, soil-type, drainage, burning by Vulgare) and Cat's-ear (Hypochoeris Aborigines and browsing by native radicata). McDougall (l989a) noted that herbivores. In some areas, the distribution regular mowing led to a marked reduction of trees at the time of settlement may have

Vol. 108 No.3 (1991) 61 represented only a temporary phase in a habitat variability. Obviously, trees should shifting mosaic, affected by burning, not be planted at densities greater than browsing and weather conditions, rather that of the original woodland. than a static pattern determined by climate or edaphic factors. Stock grazing has prevented the regen- Herbicides and fertilizers eration of trees in many grassy woodlands. Herbicides can dramatically alter the However, where trees remain, consider- composition of lowland grasslands and able regeneration may occur after stock are grassy woodlands. Herbicides are often removed. In such instances (e.g. at used to control grass growth for fire pro- Gellibrand Hill Park, and on many un- tection. However, their indiscriminate use burnt roadsides in eastern and western typically kills native plants and results in Victoria) dense regeneration may conflict vigorous stands of exotic species, such as with the desire to retain an open woodland Canary-grasses (Phalaris species), Cocks- with scattered trees, as dense trees can foot (Dactylis glomerata) and Yorkshire cause dramatic changes in the composi- Fog (Holcus lanatus), which are of greater tion of the grassy ground layer (Robertson fire hazard than the original natives 1985, McDougall I989a). Such regenera- (Stuwe 1981). However, selective tion may be controlled by burning or application of herbicides can play an grazing by wallabies. Robertson (1985) important role in controlling weeds in found that regeneration of tree seedlings lowland grasslands, including woody was severely curtailed by burning every weeds such as Briar Rose (Rosa rubi- six years, and Allen (1987) reported that ginosa) and dense infestations of exotics browsing by Swamp Wallabies controlled such as Chilean Spear-grass (Stipa tree and shrub regeneration. neesiana) and Canary-grasses (Phalaris species). Research is currently underway on the use of selective herbicides to de- Thee planting plete exotic species without adversely Lowland grasslands have, by definition, affecting native grasses (Shears 1988, few or no trees, and grassy woodlands McDougall 1989a,b, Morgan 1989). How- have an open canopy of scattered trees. ever, research has not yet been conducted Consequently, dense plantings of trees or on the susceptibility of native forbs, most shrubs will dramatically alter the original of which are considerably· rarer than the structure of these ecosystems. Tree dominant native grasses. planting has a dramatic effect on ground Most Australian plants are adapted to plants as well. Distinctive zones occur in low nutrient levels. Many exotic grasses the gaps between trees, in 'haloes' sur- show a grea.ter response to additional rounding trees, and beneath the canopies nutrients than do native grasses such as (Robertson 1985). Kangaroo Grass does Kangaroo Grass and Wallaby-grasses not compete well beneath trees, where it is (Groves et al. 1973,Fisher 1974,Culvenor replaced by Spear-grasses (Stipa species) 1981).Consequently, fertilizers give exotic and Wallaby-grasses (Danthonia species) grasses a competitive advantage over (McDougall 1989a). This effect may native plants, and may promote dense extend up to 30 m beyond the tree canopy stands of exotics which can smother and (McDougall 1989a). out-compete native plants. Exotic species Although tree planting is undesirable in may be benefited most greatly when fertil- native grasslands, it may have a positive izers are applied to disturbed soil (Hobbs role in grassy woodlands in which the and Atkins 1988).Fertilizers should never original tree cover has declined, by pro- be applied to native grasslands and grassy moting plant diversity by enhancing woodlands.

62 Victorian Nat. Rehabilitation and re-creation It is important to remember that a native Due to the ubiquitous impact of grassland is an ecosystem containing a European land use, particularly agricul- wide variety of plants and animals. A re- ture, all remnant lowland grasslands and created, mono-specific stand of Kangaroo grassy woodlands are degraded to some Grass is not a grassland ecosystem. As degree. As noted by Stuwe (1986), 'it could noted by McDougall (1989a) 'grassland be argued that pristine native grasslands restoration ... is not an alternative to on the western plains, and probably preservation of remnants: throughout lowland Victoria generally,are Notwithstanding the above criticisms, extinct: Consequently, some degree of the long-term conservation of many grass- vegetation management is required at all land ecosystems may rely upon the use of remnant sites. re-created sites to augment small and Management of degraded ecosystems . isolated natural remnants. Reconstructed invariably involves rehabilitation - the in- ecosystems may be used for ex situ plant troduction (or reintroduction) of depleted conservation and to propagate species for plants or animals to protected remnants. re-establishment in natural ecosystems. The intensity of rehabilitation may vary They may also have value for greatly, depending upon the level of past horticultural, landscape and educational and continuing degradation and the aims reasons. of management. Minimal rehabilitation may involve the introduction of rare and Summary endangered species into intact lowland Conservation oflowland grasslands and grasslands and grassy woodlands. grassy woodlands is dependent upon re- Maximal rehabilitation includes attempts servation and informed vegetation to re-create entire grassland and grassy management. Although further research woodland ecosystems from scratch. on most aspects of vegetation manage- A popular myth persists that grasslands ment is urgently required, some are 'simple' ecosystems, which are simple generalisations can be made. to create, by virtue of their seemingly Introduced stock eat palatable species simple structure (i.e. one apparent layer). (e.g. many daisies) in preference to most However, grassland structure may be ex- dominant grasses (including Kangaroo tremely complex at the small scale (e.g. Grass) and will invariably diminish the Mitchley 1988).Moreover, the complexity conservation value of grassy ecosystems. of these ecosystems arises not merely from Most importantly, stock should not their structural form, but from their flor- graze remnant grasslands and grassy istic and faunal composition and inter- woodlands that have not previously been actions. Some grassy wooolands in grazed (e.g. rail-lines and some roadsides). western Victoria that are structurally Any action that breaks the soil surface, 'simple' are among the richest vegetation including ploughing, grading, mowing and communities in the temperate world, with ripping, will degrade lowland grasslands up to 45 species in a single square metre and grassy woodlands. Fertilizers should (Lunt I99Oc).Their re-creation is, for all never be applied. intents and purposes, impossible. There is An appropriate burning program no evidence to suggest that lowland grass- appears to be the most suitable form of lands or grassy woodlands are any easier management for intact Themeda grass- to construct than are other ecosystems, lands. In general, Themeda grasslands such as rainforests, forests or heathlands. should be burnt at least every five years. Artificial grasslands and grassy woodlands Themeda grasslands should be burnt are rarely,if ever, likely to approximate the rather than mown. However, if mown, ecological integrity of natural ecosystems. native grasslands should not be mown

Vol. 108 No.3 (1991) 63 more than once every three months, and National Park'. Arthur Rylah Institute for Environmental Research, Technical Report preferably not more than once a year. Series No. 72. (Department of Conservation, Trees should not be planted in native Forests and Lands: Melbourne). grasslands. Trees should not be planted in Culvenor, R.A. (1981). Aspects of the Phosphorus Nutrition of Some Australian Native Grass grassy woodlands at densities greater than Species. (M.Agric.Sci. Thesis, laTrobe Uni- that of the original woodland. versity: Melbourne). A native grassland is an ecosystem con- Davies, R.J-P. (1982). 'The Conservation ofMl!jor Plant Associations in South Australia: (Con- taining a wide variety of plants and servation Council of South Australia: animals. A reconstructed, mono-specific Adelaide). stand of Themeda is not a grassland eco- Davis, S.D., Droop, S.J.M., Gregerson, P., Henson, L., Leon, C.J.,.VilIa-Lobos, J.L., Synge, H. and system. Attempts to re-create lowland Zantovska, J. (1986). 'Plants in Danger. What grasslands and grassy woodlands 'from Do We Know?' (International Union for Con- scratch' are supplements, not alternatives, . servation of Nature and Natural Resources: ). to the conservation ofintact remnants. Farrington, P. and Mitchell, A. (1966). The Effects of Grazing by Sheep and Rabbits in a Victorian Acknowledgements Forest. (Unpubl. Report, Soil Conservation Authority Victoria: Melbourne) .. Thanks are due to David Cheal, Paul Fensham, R.J. and Kirkpatrick, J.B. (1989). The Gullan, Keith McDougall, Bob Parsons conservation of original vegetation remnants in and John Stuwe for comments on various the Midlands, Tasmania. Papers and Proceed· ings of the Royal Society of Tasmania 123, drafts of the manuscript; to Doug Frood 229-246. for comments on Fig. I; and to Juliet Fisher, H.J. (1974). The effects of nitrogen fertil- McCallum and Fiona Young for patiently izer on a kangaroo grass (Themeda australis) grassland. Australian Journal of Experimental digitising Fig"1 from my rough maps, Agriculture and Animal Husbandry 14, using computer programs prepared by 526-532. Paul Gullan. Forests Commission Victoria (1984). Forests of Victoria. 1:1,000,000 map. (Government Printer: Melbourne). Frood, D. and Calder, M. (1987). 'Nature Conserv- References ation in Victoria. Study Report'. (Victorian Allen, G.G. (1987). Swamp Wallabies, Wa/labia National Parks Association: Melbourne). bicolor, and Their Interactions with a Regener- Gibbons, ER. and Downes, R.G. (1964). 'A Study ating Woodland at Gellibrand Hill, Victoria. of the Land in South-western Victoria'. (Soil (B.Sc.Hons. Thesis, Botany Department, Conservation Authority Victoria: Melbourne). University of Melbourne: Melbourne). Groves, R.H. (1965). Growth of Themeda Bennett, A.E (1982). The mammals of the Wools- australis tussock grassland at St. Albans, thorpe area, western Victoria: and the changes Victoria. Australian Journal of Botany 13, that have occurred since European settlement. 291-302. Victorian Naturalist 99, 229-240. Groves, R.H. (1979). The status and future of Bride, T.F. (1898). 'Letters from Victorian Australian grasslands. New Zealand Journal of Pioneers Being a Series of Papers on the Early Ecology 2,76-81. Occupation of the Colony, the Aborigines, etc'. Groves, R.H., Keraitis, K. and Moore, C.W.E. with (New Edition 1969, William Heinemann: appendix by Langer, H.J. (1973). Relative Melbourne). growth of Themeda australis and Poa labil· Chan, C.W. (1980). Natural Grasslands in Can- lardieri in pots in response to phosphorus and berra: Their Distribution, Phenology and nitrogen. Australian Journal of Botany 21, Effects of Mowing. (M.Sc. Thesis, Australian I-II. National University: Canberra). Hobbs, R.J. and Atkins, L. (1988). Effect of dis- Cheal, D. (1986). A park with a kangaroo problem. turbance and nutrient addition on native and Oryx 20, 95-99. introduced annuals in plant communities in the Cheal, D. (1988). Kangaroos in Hattah-Kulkyne Western Australian wheatbelt. Australian National Park. Victorian Naturalist 105, 155. Journal of Ecology 13, 171-179. Conley, D.N. (1984). Patterns of agriculture. In Kirkpatrick, J.B. (1986). The viability of bush in 'The Western Plains - A Natural and Social cities - ten years of change in an urban grassy History'. Eds D.Conley and C.Dennis, pp.15- woodland. Australian Journal of Botany 34, 26. (Australian Institute for Agricultural 691-708. Science: Melbourne). Kirkpatrick, J., Gilfedder, L. and Fensham, R. Coulson, G. and Norbury G. (1988). 'Ecology and (1988). 'City Parks and Cemeteries. Tasmania's Management of Western Grey Kangaroos Remnant Grasslands and Grassy Woodlands'. (Macropus fuliginosus) at Hattah.Kulkyne (Tasmanian Conservation Trust: Hobart).

64 Victorian Nat. Koopowitz, H. and Kaye, H. (1983). 'Plant Extinc- Department of Conservation, Forests and tion: A Global Crisis'. (Stone Wall Press: Lands: Melbourne). Washington). Ministry for Planning and Environment (1989). Land Conservation Council (1987). 'Report on the 'Planning Guide for Native Vegetation Reten- Mallee Area Review'. (Government Printer: tion'. (Government Printer: Melbourne). Melbourne). Mitchley, J. (1988). Control of relative abundance Land Conservation Council (1988). 'Statewide of perennials in chalk grassland in southern Assessment of Public Land Use, Victoria, England. II. Vertical canopy structure. Journal Australia'. (Government Printer: Melbourne). of Ecology 76, 341-3S0. Leigh, J.H. and Holgate, M.D. (1978). Effects of Moore, R.M. (19S9). Ecological observations on pasture availability on the composition and plant communities grazed by sheep in Aus- quality of the diet selected by sheep grazing tralia. Monographiae Biologicae 8, SOO-S 13. native, degenerate and improved pastures in the Moore, R.M. (1962). Effects of the sheep industry Upper Shoalhaven Valley, New South Wales. on Australian vegetation. In 'The Simple Fleece Australian Journal of Experimental Agriculture - Studies in the Australian Wool Industry'. Ed. and Animal Husbandry 18, 381-390. A. Barnard, pp. 170-183. (Melbourne Uni- Lodder, M., Groves, R.H. and Wittmark, B. versity Press: Melbourne). (1986). Native grasses - the missing link in Moore, R.M. (1964). Ecological effects of grazing Austlian~ landscape design. Landscape Aus· on grasslands in south-eastern Australia. Pro- tralia I, 12-19. ceedings of the IXth International Grassland Lunt, I.D. (1987). Effects of Environment, Land Congress, Brazil, pp. 429-433. Use and Competition on Vegetation Patterns in Moore, R.M. and Biddiscombe, E.F. (1964). The a Temperate Themeda triandra Forsk. Grass- effects of grazing on grasslands. In 'Grasses and land in Victoria, Australia. (B.Sc.Hons. Thesis, Grasslands'. Ed. C.Barnard, pp. 221-23S. laTrobe University: Melbourne). (Macmillan: London). Lunt, I.D. (1990a). Conservation management of Morcom, L. (1990). The Effects of Grazing by native grasslands and grassy woodlands. In Introduced Stock on the Vegetation of Terrick 'Management of Amenity and Sports T)Jrf'. Ed. Terrick State Park. (Unpublished report, De- T.Arthur and D.Aldous, pp. 93-102. (Royal Aus- partment of Conservation, Forests and Lands: tralian Institute of Parks and Recreation: Bendigo). Melbourne). Morgan, J.W. (1989). Weed control in native grass- Lunt, I.D. (1990b). Impact of an autumn fire on a lands. I. Chemical weed control strategies long-grazed Themeda triandra (Kangaroo (abstract). In 'Control of Environmental Grass) grassland: implications for management Weeds', Eds R. Adair and R. Shepherd, pp. 6-9. of invaded, remnant vegetations. Victorian (Weed Science Society of Victoria Inc. and Naturalist 107, 4S-S 1. Department of Conservation, Forests and Lunt, I.D. (1990c). Species-area curves and Lands: Melbourne). growth.form spectra for some herb-rich wood- Mueck, S.G. (1988). Vegetation Monitoring lands in western Victoria, Australia. Australian Survey October 1987 Hattah-Kulkyne National Journal of Ecology IS, ISS-161. Park. (Unpublished Report, National Parks Lunt, I.D. (in press). Rare or threatened plants of Service Victoria: Melbourne). lowland grasslands and grassy woodlands in Murphy, D.D. and Ehrlich, P.R. (1989). Conserva- Victoria. Advances in Nature Conservation I. tion biology of California's remnant native Mack, R.N. (1989). Temperate grasslands vulner- grasslands. In 'Grassland Structure and able to plant invasions: characteristics and Function: California Annual Grassland'. Eds consequences. In 'Biological Invasions: A L.E Huenneke and H.Mooney, pp.201-211. Global Perspective'. Eds J.A. Drake, H.A. (Kluwer: Dordrecht). Mooney, E diCastri, R.H. Groves, EJ. Kruger, Neave, H.M. and Tanton, M.T. (1989). The effects M. Rejmanek and M. Williamson.pp. ISS-179. of grazing by kangaroos and rabbits on the (John Wiley and Sons: Chichester). vegetation and the habitat of other fauna in the McDougall, K. (1987). 'Sites of Botanical Tidbinbilla Nature Reserve, Australian Capital Significance in the Western Region of Territory. Australian Wildlife Research 16, Melbourne'. (Melbourne Western Region 337-351. Commission Inc. and Department of Nicholson, P.H. (1981). Fire and the Australian Conservation, Forests and Lands: Melbourne). Aboriginal - an enigma. In 'Fire and the Aus- McDougall, K.L. (1989a). 'The Re-establishment tralian Biota'. Eds. A.M.GilI, R.H.Groves and of Themeda triandra (Kangaroo Grass): Impli- I.R.Noble, pp. SS- 76. (Australian Academy of cations for the Restoration of Grassland'. Science: Canberra). Arthur Rylah Institute for Environmental Paine, D.W.M. (1982). Vegetation. In 'Atlas of Research, Technical Report Series No. 89. Victoria'. Ed. J.S. Duncan, pp.49-S6. (Govern. (Department of Conservation, Forests and ment Printer: Melbourne). Lands: Melbourne). Powell, J.M. (1967). Victoria's woodland cover in McDougall, K.L. (1989b). Weed control in native 1869. A bureaucratic venture in cartography. grasslands. II. The use of atrazine and fire New Zealand Geographer 23, 106-116. (abstract). In 'Control of Environmental Robertson, D. (198S). Interrelationships Between Weeds'. Eds R.Adair and R.Shepherd, pp. lO- Kangaroos, Fire and Vegetation Dynamics at ll. (Weed Science Society of Victoria Inc. and Gellibrand Hill Park, Victoria. (Ph.D. Thesis,

Vol. 108 No.3 (1991) 6S Botany Department, University of Melbourne: Groves, pp. 393·410. (Cambridge University Melbourne). Press: Cambridge). Robinson, G.G. and Dowling, P.M. (1976). Man- Stuwe, J. (1981). Native grasses as a management agement of natural pastures on the Northern tool. Australian Parks and Recreation May, Tablelands of New South Wales - A Survey. 39-44. Australian Rangelands Journal I, 70-74. Stuwe, J. (1986). 'An Assessment of the Conserva- Ross, J.H. (1990). 'A Census of the Vascular Plants tion Status of Native Grasslands on the Western of Victoria.' 3rd.ed: (Department of Conserva- Plains, Victoria and Sites of Botanical tion, Forests and Lands: Melbourne). Significance'. Arthur Rylah Institute for Rowan, J. (1988). Land Systems of Victoria. (Un- Environmental Research Technical Report published Report, Land Conservation Council: Series No.48. (Department of Conservation, . Melbourne). Forests and Lands: Melbourne). Scarlett, N.H. and Parsons, R.F. (1982). Rare Stuwe, J. and Parsons, R.F. (1977). Themeda plants of the Victorian plains. In 'Species at australis grasslands on the Basalt Plains, Vic- Risk: Research in Australia'. Eds R.H. Groves toria: floristics and management effects. and W.D.L. Ride, pp.89-105. (Australian Australian Journal of Ecology 2, 467·476. Academy of Science: Canberra). van Rees, H. (1982). The diet of free-ranging cattle Scott, A.w. and Whalley, R.D.B. (1984). The influ- on the Bogong High Plains, Victoria. Australian ence of intensive sheep grazing on genotypic Rangelands Journal 4, 29-33. differentiation in Danthonia linkii, D. van Rees, H. and Holmes, J.H.G. (1986). The richardsonii and D. racemosa on the New botanical composition of the diet of free- England Tablelands. Australian Journal of ranging cattle on an alpine range in Australia. Ecology 9,419-429. Journal of Range Management 39. 392·395. Seebeck, J.H. (1984). Mammals of the plains or, Wells, T.C.E. (1969). Botanical aspects of conserv- where have all the wombats gone? In 'The ation management of chalk grasslands. Western Plains - A Natural and Social History'. Biological Conservation 2, 36-44. Eds D. Conley and C. Dennis, pp.39-53. (Aus- Whalley, R.D.B., Robinson, G.G. and Taylor, J.A. tralian Institute for Agricultural Science: (1978). General effects of management and Melbourne). grazing by domestic livestock on the rangelands Shears, I.G. (1988). Phytotoxic Effects of of the Northern Tablelands of New South FIuazifop-butyl, Sethoxydim and Ethofumasate Wales. Australian Rangelands Journal I, on Native and Exotic Grass Species. (Thesis, 174-190. Victorian College of Agriculture and Horti- Williams, M. and Goodwins, D. (1988). Conserva· culture, Burnley Campus: Melbourne). tion of biological diversity on the FIeurieu Sim, I. and Urwin, N. (eds.) (1984). 'The Natural Peninsula. South Australian Naturalist 63, Grasslands of the Liverpool Plains, New South 24-39. , Wales'. (Department of Environment and Woodgate, P. and Black, P. (1988). 'Forest Cover Planning: Sydney). Changes in Victoria 1869·1987'. (Department Specht, R.L. (1972). 'The Vegetation of South Aus· of Conservation, Forests and Lands: tralia'. 2nd ed. (Government Printer: Adelaide). Melbourne). Specht, R.L. (1981). Conservation of vegetation types. In 'Australian Vegetation.' Ed R.H.

66 Victorian Nat.