Threatened species & ecological communities

Epacris stuartii Recovery Plan

David Keith, Revised by Mick Ilowski
Department of Primary Industries, Water and Environment
Environment Australia, November, 1999
ISBN 0 7246 4290 0

Contents

Figures

SUMMARY

CURRENT SPECIES STATUS:

Epacris stuartii is classified as Endangered under both Tasmanian state and Commonwealth endangered species legislation. There is only one wild population of about 850 mature plants. The population has declined by approximately 20% between the beginning of 1994 and March 1996.

HABITAT REQUIREMENTS AND LIMITING FACTORS:

Epacris stuartii is restricted to 0.3 ha of heathland on an exposed headland near Southport, south-east Tasmania. Plants occur in shallow soil or in crevices on outcrops of dolerite. Major threats are: disease epidemic caused by Phytophthora cinnamomi; fire regimes involving high frequency fires and/or fires followed by drought; and extreme storm events.

RECOVERY OBJECTIVES:

To minimise the chance of extinction of Epacris stuartii by:

RECOVERY CRITERIA:

RECOVERY ACTIONS:

ESTIMATED COST OF RECOVERY:

Costs are in 1996 dollars. Costs for the 6 years 2000-2005 are identical and quoted on a per annum basis(*).

Estimated cost of recovery
Year:            
Action Priority 1 2 3 4 5 - 10* Total
1.1 1 7190 2710 2710 2710 2710 18030
1.2 Contingency actions - see main text for indicative costs
2.1 Costs included within Action 1.1
2.2 1 2550 800 800 800 800 5750
2.3a 1 8120         8120
2.3b 2 20627         20627
2.4 Contingency action - see main text for indicative costs
3.1 1 16024 10988       27012
3.2 2 11994         11994
3.3 1 5052         5052
4.1 1-Feb   26575   48700   75275
4.2 Contingency action - see main text for indicative costs
Totals   71557 41073 52210 3510 3510 171860

1. INTRODUCTION

1.1 Description and Classification of Species

Epacris stuartii belongs to a genus of about 40 species endemic to south-eastern Australia and New Zealand. Epacris belongs the family Epacridaceae which has a Gwondwanan distribution including Australia, south-east Pacific Islands, South America, Malaysia and Indo-China. The Epacridaceae includes a group of genera (tribe Epacrideae), including Epacris, Sprengelia, Richea, Andersonia and Dracophyllum, distinguished by their dry dehiscent fruits bearing multitudinous tiny seeds. The other tribe within the family (Styphelieae) includes genera such as Leucopogon, Styphelia, Trochocarpa, Acrotriche and Cyathodes, characterised by indehiscent fleshy fruits bearing few seeds. The Epacridaceae is placed in the Order Ericales, along with one other family, Ericaceae, which is distributed principally in forests and heathlands of the northern hemisphere.

Epacris stuartii is an erect or semi-prostrate multi-stemmed shrub growing up to 1 m tall. Its branches are robust, bearing ovate-cordate (heart-shaped) leaves tending to ovate (oval-shaped) on young branches, 4-7 mm long and 3-5 mm wide with short stalks (<1 mm long). The leaves have a thick glossy cuticle, a pungent mucronate apex and three parallel veins on the lower surface. The flowers appearing in late winter - early spring are white, solitary in the leaf axils, subsessile and crowded along the upper parts of the branches. The style and anthers are prominently exserted from the corolla tube which is 3-4 mm long and has five lobes 3.5 - 5 mm long. Fruits are capsules up to 2 mm long and enclosed until dehiscence within imbricate whorls of sepals and bracts. The seeds are tiny and numerous within the fruits (Curtis 1963, Crowden and Menadue 1990).

Epacris stuartii belongs to a group of about 10 closely related species known as the 'Epacris tasmanica complex' (Crowden and Menadue 1990). This group is endemic to Tasmania and includes several other threatened species, including E. barbata, E. virgata, E. apsleyensis, E. exserta and E. glabella. Crowden and Menadue (1990) concluded from a morphometric analysis based on selected leaf and floral characters that E. stuartii was distinct from other taxa in the group. The analysis showed the closest relatives of E. stuartii to be E. virgata sens. str., E. sp. aff. exserta (Mt Cameron) and E. tasmanica (Group A), a small-flowered taxon more similar to E. virgata than E. tasmanica sens. str. which was placed in Group B (large flowers). Epacris stuartii was distinguished from these taxa in the analysis by its longer leaves and corolla lobes (petals). Several characters that were not included in Crowden and Menadue's (1990) analysis are likely to reaffirm the distinctness of E. stuartii from other taxa in the group. These include the thickness and luster of its leaves (thicker and more glossy in E. stuartii than other taxa in Group A), the length of its inflorescences (less elongated in E. stuartii), the width of its basal and ultimate branches (more robust in E. stuartii), and its habit (more usually semi-prostrate in E. stuartii cf. always erect in other taxa). The taxonomic status of some other members of the Epcaris tasmanica complex remains to be formally resolved (Crowden and Menadue 1990).

1.2 Conservation Status

Epacris stuartii is currently classified as an Endangered species under both Tasmanian state and Commonwealth endangered species legislation. It had previously been regarded as Vulnerable (Leigh et al. 1981, Briggs and Leigh 1988). The change in status appears to have resulted from a reappraisal of available information on its distribution and apparent threats, rather than evidence of recent decline in population size or range.

1.3 Distribution and Abundance

1.3.1 Distribution

Epacris stuartii has only ever been known from a single locality on Southport Bluff, about 6 km south-east of Southport township in far south-eastern Tasmania (Fig. 1, Appendix 1). The population occupies an area of approximately 0.3 ha over a range of about 300 m.

Searches of apparently similar habitat have been carried out by Dr R. K Crowden (University of Tasmania), Mrs K. Geeves (Society for Growing Australian Plants) and recently by Dr P. J. C. Barker (Forestry Tasmania). Searches carried out by Barker were part of a survey project funded by the Endangered Species Program (Project No. 508) and will be reported later in 1996. The searches have examined rocky dolerite coastline in the vicinity of Southport Bluff (Fig. 1). All searches were unsuccessful. Some similar habitats in the area, such as clifftop heath on South Bruny Island support Epacris myrtifolia, rather than E. stuartii, or no Epacris species at all. Therefore it seems unlikely that other populations of E. stuartii exist. However, areas of apparently similar habitat that remain to be searched occur along the inaccessible coastline between Cockle Creek and South East Cape.

The population of Epacris stuartii is disjunct from populations of related taxa in the E. tasmanica complex. The closest known populations belong to the taxon described as E. tasmanica (Group A) by Crowden and Menadue (1990) and included by them within a broadened concept of E. virgata. These populations occur 30 km to the north-east near Adventure Bay on South Bruny Island and 45 km to the north between Snug and Kettering and are separated from E. stuartii by wide bodies of water (D'Entrecasteaux Channel and the Huon River estuary, respectively). Other taxa in the E. tasmanica complex occur on the east Tasmanian coast north from Hobart and at scattered locations in northern and western lowland Tasmania (Crowden and Menadue 1990). The geographic isolation of E. stuartii from related taxa, its very localised seed dispersal (Section 1.5.4) and the presumed small home range of its invertebrate pollinators (Section 1.5.2) may be major factors involved in its speciation.

Distribution and Abundance

Figure 1: Distribution of Epacris stuartii (*) showing areas of coastline searched unsuccessfully for new populations (circled). The southern shores of Port Esperance, the coastline between Dover and Huon Point and at Fluted Cape (Bruny Island) were also searched unsuccessfully.

1.3.2 Population Size and Structure

In 1995 the population of Epacris stuartii at Southport Bluff was estimated to contain a total of ca. 1,000 plants, of which ca. 85% were estimated to be reproductively mature (Fig. 2). The estimate was derived from mean population density (3 plants per square metre) multiplied by the area of occurrence (300 m2), which was mapped by foot traverse. The estimate was validated by an exhaustive census of plants within an area comprising about one-third of the total area occupied by the population. This census yielded 400 individuals.

The population of E. stuartii comprised plants that varied in size and reproductive status. Three size classes were defined for established plants by summing the length of all their basal stems (small: <50 cm, medium: 50-100 cm and large: >100 cm). Thus individuals were placed in the largest size class either if they had one or two long stems (erect growth form) or many short stems (semi-prostrate growth form). An additional size class was defined to include seedlings, recognised by their short (<10 cm) slender stems and small leaf size. In February 1995 less than 1% of the total population was represented by seedlings These were presumed to have emerged in the previous year. The population structure varied between substrates (see Section 1.4.2). On soil, more than 41% of plants were large, compared to 18% on intermediate substrates and 5% on rock substrates (Fig. 2). Conversely only 26% of plants on soil substrates were small compared to 36% on intermediate substrates and 72% on rock substrates (Fig. 2). The percentage of reproductive plants varied from 55-78% in the small size class, to 75-95% in the medium class, to 94-100% in the large size class (Fig. 2).

Figure 2: Population structure in three microhabitat types

Figure 2: Population structure in three microhabitat types.

1.4 Habitat

1.4.1 Physical Environment

Epacris stuartii occurs on an exposed dolerite headland at an elevation of 5 to 19 m above sea level. The headland is semi-circular in shape and has a flat top inclined gently to the west and terminating in a small interrupted cliffline around its northern, eastern and southern edges. Epacris stuartii is associated with seaward slopes and rocky outcrops. Plants are scattered mostly around eastern and southern aspects in numerous small subcatchments draining different parts of the headland. Despite the range of aspects, the entire population is exposed to onshore winds varying from slight to gale force.

1.4.2 Substrate Microhabitat

There is small-scale variation in the substrate microhabitat of E. stuartii. The majority of the population occurs in shallow, well-drained, dark grey loam - sandy loam (Fig. 2). In this microhabitat soil depth varies up to ca. 20 cm and a scattering of dolerite rocks may protrude through the soil surface. A smaller, but significant proportion of the population occurs on massive rocky outcrops (Fig. 2), either in crevices with no visible soil, or in shallow depressions with up to 1-2 cm of soil or moss, presumably with roots penetrating cracks in the subsurface rock. On average, plants on rock substrates are smaller than those in shallow soil substrates (Fig. 2). Plants also grow in sites that may be regarded as intermediate microhabitats.

1.4.3 Vegetation

Epacris stuartii occurs exclusively in heathland. The structure and composition of co-occurring vegetation varies with substrate microhabitat and fire history. On shallow soil substrates the heath may be dominated by a dense shrub stratum with Leptospermum scoparium, Acacia verticellata, Westringia brevifolia and Banksia marginata up to ca. 1 m tall. Epacris stuartii occurs mainly in gaps of the canopy of these larger shrubs, though some individuals occur beneath the canopy and others protrude through it. Smaller shrubs include Pultenaea dentata, Epacris impressa, Bossiaea prostrata and Astroloma humifusum. Groundcover species include Stylidium graminifolium, Gonocarpus tetragynus, Helichrysum scorpioides, Viola hederacea, Deyeuxia densa, Microlaena stipoides, Danthonia pilosa, Poa poiformis and Lepidosperma sp.. In recently burnt areas the cover of large shrubs is reduced and there is an almost continuous groundcover dominated by Lomandra longifolia and Deyeuxia densa with various other grasses and sedges. On rocky substrates the cover of vegetation is sparse, with scattered individuals of Westringia brevifolia, Leptospermum scoparium, Poa poiformis, Baumea juncea, Selliera radicans and the introduced grass Aira caryophyllea.

1.4.4 Fire History

The fire history of Southport Bluff and surrounds was established by examining fire reports prepared by the Parks and Wildlife Service, gathering observations during field inspections and by interviewing relevant personnel. Fire records for the George III Historic Site and the adjacent Southport Lagoon Wildlife Sanctuary extend back to 1978. Two fires have apparently burnt the population of Epacris stuartii on Southport Bluff between 1978 and 1996. The first fire occurred on 3rd February 1981. A detailed fire report (Skinner 1981) and recollections by Ranger R. R. Donnelly, who inspected the Bluff on 4th February 1981, both indicate that the fire was intense and that it burnt the entire population of E. stuartii. This description of the 1981 fire is supported by the remains of long dead shrubs, evidently killed in the 1981 fire, that are distributed around the cliffline and on the cliff face itself around the entire headland (pers. obs., 1995). It therefore seems reasonable to assume that virtually all E. stuartii plants were burnt in a high intensity fire, though it is impossible to tell whether a few individuals escaped the 1981 fire in small patches.

The second fire occurred on 2nd April 1994 (Bradley 1994). This fire apparently reached moderate intensity on the flat summit of the Bluff because the foliage of shrubs was entirely consumed (pers. obs., February 1995). Around the edges of the Bluff the fire was patchy. Scorched foliage remaining on plants around the cliffline (pers. obs.) suggest that the fire was generally less intense in these areas. Unburnt patches up to 50 m2 in size were scattered around the north, east and south edges of the Bluff. The spatial pattern of burnt areas appeared to be unrelated to fuel density and rocky outcrops. Both microhabitats of E. stuartii (dense heath on shallow soil and open heath on rocky outcrops) each contained a mosaic of burnt and unburnt patches. On some of the rocky outcrops the fire had spotted between individual shrubs, some of which were quite small and separated by distances of a few metres.

Both the 1981 and the 1994 fires started several kilometres to the west between Southport Lagoon and the Ida Bay Railway. There is a very high frequency of ignitions in the vicinity of Southport Lagoon, though many of these occur to the west and south of the lagoon itself (Parks and Wildlife Service fire records). The causes include arson, unattended campfires, and activities connected with the operation of the railway and quarries. Although these ignitions have to date resulted in only a moderate frequency of fires at Southport Bluff, there is a potential for a more frequent fire regime which may threaten the population of Epacris stuartii (see below).

1.5 Life History and Ecology

As part of the Research Plan (Action 1) a census was established to examine the population dynamics of the E. stuartii. A sample of established plants stratified by plant size, substrate microhabitat, fire history and shading was permanently tagged with fire-resistant metal tags. Censuses were carried out in February 1995 and January 1996 to measure rates of survival, growth and reproduction. For each plant the following data were recorded: alive or dead; substrate microhabitat (soil, rock or intermediate); shading by canopies of adjacent plant (unshaded, shaded, intermediate); effect of 1994 fire (entirely burnt, partially burnt, entirely unburnt); number and lengths of basal stems; and number of infructescences produced in the current season. It was assumed that plants retaining dead foliage in February 1995 had died during 1994.

1.5.1 Survival and Growth of Established Plants

Epacris stuartii has a low background rateof mortality (<1% per annum) and is therefore likely to be long-lived. The life span of individual shrubs is unknown, but probably in the order of decades.

Mortality during 1994 -1996 occurred in two major episodes, April 1994 and February 1996, in addition to the low level of background mortality. The first episode of mortality was related to a fire and was essentially confined to plants on rock substrates. Almost all burnt plants on soil substrates survived and resprouted. The fire caused a decline in the population of established plants by 12%.

The second episode of mortality occurred suddenly 3 weeks after a severe storm which was followed by warm weather. This resulted in a further decline in the population of standing plants of at least 15% and up to 50%, depending on the fate of surviving plants that suffered varying levels of leaf tissue death. Several other shrub species at the site showed a similar response at the same time. These were Leptospermum scoparium, Westringia brevifolia, Banksia marginata, Astroloma humifusum, Pultenaea dentata, Leucopogon parviflorus and Lomatia tinctoria. Tests of soil and root samples failed to detect evidence of Phytophthora cinnamomi. It seems likely that plant mortality and dieback were a response to extreme salt levels deposited by high winds and seas during a storm event of rare and extreme severity. This interpretation is supported by observations at the same time of similar responses in other coastal heathlands of south-east Tasmania (e.g. on Tasman Peninsula and South Bruny Island).

Among the few plant deaths not associated with either episode of mortality, the symptoms and pattern of mortality were consistent with desiccation being the cause. Overall the population of established plants has declined by at least 20% between early 1994 and March 1996.

1.5.2 Flowering Phenology, Pollination and the Breeding System

In 1995 floral buds were initiated in January and February on new growth that emerged in the previous spring. Buds on five inflorescences on a sample of 36 plants were tagged and followed between March 1995 and February 1996. The first flowers were recorded early in August (Fig. 3), though scattered individuals were observed flowering, mainly on north-facing slopes, as early as late May. Flowering was complete by late October. Fruit dehiscence began late in December and was complete by early February 1996. The phenology of flowering and fruiting observed in 1995-96 was consistent with anecdotal observations (R. Crowden, pers. comm., K. Geeves pers. comm.) and herbarium records from previous years.

The pollinators of Epacris stuartii are large adult carrion flies (family Calliphoridae). Two species of flies were directly observed pollinating E. stuartii on several days during August and September 1995: Calliphora sp. ('metallic green abdomen' group) and Calliphora hilli (Dr P. B. McQuillan, pers. comm., Department of Geography and Environmental Studies, University of Tasmania). Calliphora is a widespread genus of flies in Tasmania (P. McQuillan, pers. comm.). Epacris stuartii was apparently the only local pollen source exploited by Calliphora spp. at this time of year and seems likely to be their first floral food source after emergence as adults. After flowering of E. stuartii was complete flies were observed visiting flowers of Westringia brevifolia, Acacia verticillata and Leptospermum scoparium which flowered in an overlapping sequence on the site through the summer months into Autumn.

Figure3 Reproductive phenology of Epacris stuartiiin 1995-96

Figure 3: Reproductive phenology of Epacris stuartii in 1995-96.

It seems unlikely that outcrossing within the E. stuartii population would be limited by pollinator movements because individual flies in genera such as Calliphora are capable of travelling several hundred metres in a single flight (P. McQuillan, pers. comm.).

Experimental investigations have failed to detect any limitation in seed set due to self pollination or due to limited availability of pollinators. Application of self or cross pollen and exclusion of pollinators had no significant effect on the proportion of fruit set. Even though the present results suggest that E. stuartii has a self-compatible breeding system, the slightly higher fruit set on cross-pollinated plants relative to self-pollinated plants suggests that the species may be preferentially outcrossing.

Figure 4 Fruit production of Epacris stuartii in 1995 in relation to plant size, substrate and fire history

Figure 4: Fruit production of Epacris stuartii in 1995 in relation to plant size, substrate and fire history.

1.5.3 Fruit Production

In 1995-96 fruit production depended on plant size, fire history and shading by the canopies of neighbouring plants, but was independent of substrate. A similar pattern was

evident in 1994-95. Large plants produced more fruit than those in the small and medium size classes in the unburnt area (Fig. 4). Plants of comparable size produced similar numbers of fruits on rock and soil substrates (Fig. 4). However, overall fruit production was greater on soil substrates than on rock because the latter habitat supported very few plants in the large size class (Fig. 1). In the burnt area, fruit production was substantially reduced compared with the unburnt area (Fig. 4).

Overall about 30% of buds initiated in early 1995 produced viable fruit in early 1996. The causes of death of developing buds, flowers and fruits varied over time and in relation to fire history and shading by the canopies of neighbouring plants. The major cause of bud and fruit loss was spontaneous abortion. In the soil substrate spontaneous abortion of buds and developing fruits was greatest amongst those plants that were most shaded by the canopies of neighbouring plants. Less than 5% of buds produced viable fruits in plants that were entirely beneath the canopies of neighbouring plants (Fig. 5a) compared to ca. 60% in plants that were free from the canopies of neighbouring plants (Fig. 5c) and ca. 40% in plants that were partially beneath the canopies of neighbouring plants (Fig. 5b). On rock substrates and in the burnt area there was minimal shading by neighbouring plants, however, there were greater fruit losses to due wallaby grazing than in plants in soil in the unburnt area (Fig. 5d, e). Fruit losses caused by wallaby grazing were greatest in the burnt area where less than 10% of buds produced viable fruits (Fig. 5e). Predation by beetle larvae and damage due to physical agents such as wind and mammal trampling were minor causes of fruit loss (Fig. 5).

Figure 5 The fate of developing Epacris stuartii buds and fruits

Figure 5: The fate of developing Epacris stuartii buds and fruits (viable, aborted, destroyed by fruit predators, destroyed by grazers or destroyed by physical agents).

(a) unshaded by neighbouring plants in unburnt vegetation on soil substrate, (b) partially shaded by neighbouring plants in unburnt vegetation on soil substrate, (c) fully shaded by neighbouring plants in unburnt vegetation on soil substrate, (d) unshaded by neighbouring plants in unburnt vegetation on rock substrate, and (e) unshaded by neighbouring plants in burnt vegetation on soil substrate.

1.5.4 Seed Dispersal

Seed release occurs during January and February (Fig. 4). Seeds are dispersed passively and lack structures that may assist dispersal by wind or animals. Quantitative data on seed dispersal are lacking. However, very few seeds are likely to be dispersed more than a few metres from their parent plant. The clustering of seedlings within a metre of adult plants supports the inference that seed dispersal is very localised.

1.5.5 Seed Dormancy and Dynamics of the Seed Bank

Seeds of Epacris stuartii are dormant, although a small fraction of each year's seed crop may be non dormant. Experimental investigation of the seed dormancy mechanism suggests that darkness, pre-heating and smoke derivatives may each have a role in breaking seed dormancy (Fig. 6). Low levels of germination in both light and dark where seeds were not treated with either heat or smoke derivatives suggests that E. stuartii has fire-related seed dormancy and a relatively small non-dormant seed fraction. The response to darkness suggests that seeds buried in the soil are more likely to germinate than those lying on or near the soil surface. Germination is therefore likely to be curtailed until seed burial occurs and this is likely to influence the fate of emerging seedlings. Seedlings emerging from sites relatively deep in the soil profile are more likely to survive dry weather conditions during their establishment phase than seedlings emerging from seeds on or near the soil surface (Bell et al. 1995).

Figure 6 Germination of Epacris stuartii seed under combinations of light, pre-heating to 90C for 10 minutes and treatment with smoked water and smoked filter paper. Data indicate preliminary results of a continuing experiment

Figure 6: Germination of Epacris stuartii seed under combinations of light, pre-heating to 90C for 10 minutes and treatment with smoked water and smoked filter paper. Data indicate preliminary results of a continuing experiment.

The longevity of seeds is unknown. However, the timing of seedling emergence suggests that E. stuartii is likely to have a persistent soil seed bank. Seedlings that emerged in 1995 were likely to have been derived from seeds released almost two years earlier in the fruiting season before the 1994 fire. Adult plants in the vicinity of seedlings were either killed by the fire or resprouted and failed to produce seeds in the only reproductive season before the seedlings emerged. The clustering of seedlings around their presumed parent plants suggests that seeds were unlikely to have been dispersed from plants in unburnt areas.

1.5.6 Seedling Recruitment and Establishment

The total number of seedlings emerging in 1994 and 1995 was small. The 412 recorded seedlings occurred in an area occupied by about 400 mature plants. Less than half of these seedlings survived beyond six months after emergence and seedling survival was higher on soil than on rock substrates (Fig. 7). The time it takes for seedlings to become mature and large enough to survive fire (in soil substrates) is unknown but likely to be several years.

Seedling emergence was almost entirely confined to the burnt area, more abundant on rock than on soil substrates and occurred almost exclusively in 1995 (Fig. 7). This spatial pattern of seedlings in relation to the burnt area is consistent with the existence of fire-related seed dormancy mechanisms and the production of a small fraction of non-dormant seeds. A similar relationship between fire and seedling emergence has been observed in other Epacris species (Keith 1991).

Figure 7 Emergence of Epacris stuartii seedlings in (a) 1994 and (b) 1995 in areas of different fire history on rock and soil substrates

Figure 7: Emergence of Epacris stuartii seedlings in (a) 1994 and (b) 1995 in areas of different fire history on rock and soil substrates. Solid bars indicate numbers of seedlings surviving at 6 months after emergence. Hatched bars indicate numbers of seedlings that failed to survive their first 6 months after emergence.

Seedling emergence was confined to the spring season in both 1994 and 1995, possibly reflecting the favourability of warmer, moist conditions for seed germination. The emergence of more seedlings in the second year after a fire than in the first year is unusual among heathland shrubs and may be in response to low soil moisture in the first post-fire year. Substantially more rainfall occurred in the spring and summer of 1995 than in the comparable seasons of 1994.

1.5.7. Propagation

Propagation of Epacris stuartii has been carried out successfully from cuttings collected in May 1995. The cuttings consisted of new growth on older wood. Three methods of cutting treatment were tested: basal dipping in undiluted ESI ROOT (ER); basal dipping in professional strength Clonex; and immersion in 0.35% ESI ROOT solution followed by basal dipping in professional strength Clonex. The two basal dipping methods were most successful in striking cuttings, yielding greater than 80% strike and survival at 18 weeks after propagation (Fig. 8). Dunking in diluted ESI ROOT and subsequent basal dipping in Clonex was a significantly less successful propagation technique, yielding 30% strike and survival rate (Fig. 8).

It appears possible to propagate Epacris stuartii from seed (section 1.5.5), although germinants have not yet been grown on beyond cotyledon stage. If the success of germination as a propagation technique is confirmed, it may prove to be a more cost-effective method of establishing and maintaining an ex situ living collection than propagation by cuttings.

Figure 8 Strike and survival rate of Epacris stuartii cuttings collected and propagated in May 1995

Figure 8: Strike and survival rate of Epacris stuartii cuttings collected and propagated in May 1995. Solid bars indicate cuttings that struck and survived to 18 weeks. Grey shaded bars indicate cuttings that failed to strike or survive. A total of 146 cuttings were tested from 14 genets in the wild population (data courtesy of M. Fountain, Royal Tasmanian Botanic Gardens).

1.6 Reasons for Listing

Epacris stuartii is listed as an endangered species by the Australian and New Zealand Environment and Conservation Council and in Schedule 1, Part 1 of the Commonwealth Endangered Specie Protection Act (1992). It is also gazetted as an endangered species under the Tasmanian Rare and Vulnerable and Endangered Species Protection Act (1996). The principal reasons for its endangered status include: its extremely restricted natural geographic range; its small and recently declining population size; imminent threats posed by disease epidemic, certain fire regimes and extreme weather events; and its limited capacity for regeneration.

Epacris stuartii is known only from an area of 0.3 ha and has a total population of about 850 mature plants, which has declined by 20% over two years since early 1994. Seedling recruitment over this period has been very localised and insufficient in magnitude to maintain population stability.

Epacris stuartii is known to be susceptible to the fungal pathogen, Phytophthora cinnamomi (Barker and Wardlaw 1995), which is associated with disease epidemics causing extreme declines in populations of related Epacris species on the east coast of Tasmania (Keith 1995, unpubl. data). Phytophthora cinnamomi has been recorded along the walking track within 1 km of the E. stuartii population and, given current usage of the track by bushwalkers, is highly likely to invade the site in the near future.

Epacris stuartii may be threatened by fire regimes that comprise fires repeatedly followed by drought conditions and fires recurring at high frequency. A fire caused the Epacris stuartii population to decline by about 15% in 1994. The reasons for this decline appear to be associated with weather conditions in the year after the fire, which were apparently too dry to support seedling recruitment. Thus, post-fire seedling recruitment was insufficient to offset the number of established plants killed by fire. It is also likely that Epacris stuartii is adversely affected by high frequency fire regimes, since these have been implicated in population declines of a wide range of woody shrub species (Keith 1996). Although the population has not experienced high fire frequencies in recent times, the potential for this to occur in future is underscored by the extremely high frequency of ignitions associated with recreational use of the adjacent Southport Lagoon Wildlife Sanctuary and surrounding lands.

Epacris stuartii may also be threatened by severe storms though physical damage and excessive salt loads. Such an event in February 1996 apparently caused the death of numerous plants and severe defoliation in about two-thirds of the population. The frequency of severe storm events associated with this level of mortality and defoliation is unknown. When combined with other threats (disease epidemic and adverse fire regimes) and the limited capacity for regeneration in the E. stuartii population, the impact of severe storms may be substantial.

1.7 Existing Conservation Measures

The only population of Epacris stuartii is protected within a reserve of secure tenure, George III Historic Site. No active management is carried out on the reserve other than the monitoring of fires. A management plan is currently in preparation for the adjacent Southport Lagoon Wildlife Sanctuary. The plan will address issues such as recreational usage, visitor access and fire management that are relevant to conservation of E. stuartii.

Access to the site is currently by foot along 5 km of undeveloped walking track or by boat and foot from nearby beaches. Vehicular access to the northern side of Southport Lagoon and thence George III Historic Site was closed in 1994 and the prohibition of vehicular access is likely to be continued when the management plan is adopted. Signs direct walkers from the Ida Bay Railway terminus at Deep Hole to George III Historic Site. The walk is currently promoted in an ad hoc manner by the Ida Bay Railway Company and the Lune River Youth Hostel. The existence and significance of E. stuartii is not publicised.

A monitoring program based on a sample of 250 plants has been established in 1995 as part of the Research Plan and is proposed to continue in the strategy for conservation.

A small ex situ living collection of E. stuartii was established in 1995 and is held at the Royal Tasmanian Botanic Gardens. The collection consists of approximately 90 individuals raised from cuttings collected from 13 plants in the wild population. The ex situ collection is unlikely to be a very comprehensive representation of genetic variability in the wild population and will require expansion if it is to function as an adequate safety net for the wild population.

1.8 Strategy for Conservation

This Recovery Plan will run for a term of 10 years from 1996 to 2005 inclusive. The major threat to E. stuartii is a disease epidemic caused by Phytophthora cinnamomi, which is expected to infect the population during the term of this Plan. Other risk factors such as fire regimes and the timing of critical weather events (drought and storm) may contribute to the overall threat at various times. Three strategies for conservation of E. stuartii will be implemented concurrently to address these threats:

(i) preventative measures to minimise the chance of invasion by P. cinnamomi through access management, monitoring and early detection;

(ii) remedial measures to restore the wild population or limit its decline through habitat management and disease treatment; and

(iii) a safety net to allow reintroduction, should the wild population become extinct.


2. Objectives and Criteria for recovery

The overall objective of this Recovery Plan is to minimise probability of extinction of Epacris stuartii in the wild. This is unlikely to entail downlisting the species' status from Endangered because of its naturally restricted range and the limited area of suitable habitat.

2.1 Specific Objectives

1. To maintain the wild population of E. stuartii above a minimum threshold size of 500 mature individuals.

2. To minimise the chance of invasion of George III Historic Site by Phytophthora cinnamomi and mitigate its impact on vegetation should infection occur.

3. To develop management techniques to restore the wild population of E. stuartii should it fall below the minimum threshold size.

4. To establish and maintain an ex situ collection of E. stuartii that represents, to an appropriate level, the variation within the wild population for the purpose of reintroduction should E. stuartii become extinct in the wild.

2.2 Criteria

1. The population size is maintained above 500 mature individuals in perpetuity.

2. Southport Bluff remains free of infections of Phytophthora cinnamomi in perpetuity. Infected vegetation in the vicinity is treated immediately, in the event that P. cinnamomi invades the site.

3. Management techniques to increase the size of the wild population of E. stuartii are developed and implemented if the population falls below 500 mature individuals.

4. A representative ex situ living collection of E. stuartii is established and maintained, and used as a source of material for reintroduction if the wild population becomes extinct.


3. RECOVERY ACTIONS

Implementation of the Recovery program will be overseen by the Recovery Team consisting of a botanist from Parks and Wildlife Service, a land manager from Parks and Wildlife Service (Esperance District), an officer from the Royal Tasmanian Botanical Gardens, an officer from the Endangered Species Unit (ANCA), the Tasmanian representative of the Tasmanian Conservation Trust and representatives of local business and/or community interests as appropriate. The text below describes the recovery actions, justifies their need and provides estimated costs in 1996 dollars.

3.1 Maintenance of population above threshold size.

ACTION 1.1 Population Monitoring

An up to date knowledge of the status of the population is an essential basis for future decisions for management of fire, access, etc. Prior to 1995 the population had not been visited regularly or assessed in a systematic manner. In 1995 a monitoring program was established as part of the implementation of the Research Plan. It is essential that monitoring continue so that population declines may be detected and their cause be identified as early as possible.

The existing monitoring program is based on a sample of 250 established plants. The plants are marked permanently with numbered brass tags on stainless steel stakes in an area south of George III Monument. Monitoring will continue to be carried out annually in January and will involve a census of survival and fruit production of existing plants and a search for new seedlings, which will be tagged. Fires occurring in the preceding year will be mapped on a detailed site plan with reference to Departmental fire records. It is essential that monitoring occur in January of each year when fruits are mature. Annual monitoring will continue for 10 years under this Plan. In addition, it will be necessary to carry out monitoring on a monthly basis in 1996 to assess recovery of individual plants suffering dieback, apparently as a consequence of the storm in February 1996. The fate of these plants, which represent a high proportion of the total population, is critical to future management options.

Action 1.1 costs
Year 1 2 3 4 5 6 7 8 9 10
                     
Supervision (1 day@$350) $700 $350 $350 $350 $350 $350 $350 $350 $350 $350
Vehicle (1wk@$180)* $900 $180 $180 $180 $180 $180 $180 $180 $180 $180
Office equipt (1wk@$96) $480 $96 $96 $96 $96 $96 $96 $96 $96 $96
Laboratory tests(1hr@$50) $50 $50 $50 $50 $50 $50 $50 $50 $50 $50
Salary (2wks @ $775)* $3,875 $1,550 $1,550 $1,550 $1,550 $1,550 $1,550 $1,550 $1,550 $1,550
On costs (28%) $1,085 $434 $434 $434 $434 $434 $434 $434 $434 $434
Materials & Travel $150 $50 $50 $50 $50 $50 $50 $50 $50 $50
                     
Total $7,190 $2,710 $2,710 $2,710 $2,710 $2,710 $2,710 $2,710 $2,710 $2,710

*In year 1 a project officer, vehicle and office accommodation is required for 5 weeks.

ACTION 1.2 Habitat management to restore or avoid decline in wild population.

Management actions directed at restoring the size of the population will be required if substantial population declines are detected in the monitoring program. A decline in the number of plants in the monitoring sample from 250 to 180 will be taken as an indication that the total population is approaching the critical management threshold of 500 mature plants. This will be verified by inspection of the entire population.

In the event that the population does decline to threshold size, appropriate management actions to encourage seedling recruitment will be considered. These potentially include implementation of regeneration fires, protection from wildfire, smoke treatment, water supplementation and fencing to exclude macropod grazers. At present there is insufficient knowledge to determine the conditions under which each of these actions should be implemented Research is required to determine the details of implementation (Action 3.3). With the exception of fire protection, none of the management actions listed should be implemented without further research.

The timing of population management actions cannot be forecast, since it is dependent upon when and if the wild population declines to the threshold size. Indicative costs of implementing actions are provided for one year as a guide for possible funding requirements. The Recovery Team may or may not submit funding applications to ANCA for these or related actions during the term of this Plan, depending on the status of the wild population.

(a) Fire protection: May be required to prevent population decline due to high frequency fire. The most feasible management option for fire protection is to create a fire break near the western boundary of the Historic Site and to burn back in a westerly direction. Costs include labour for planning, slashing, burning and post-fire patrol, consumables and boat hire. Repeat burns may be required at appropriate intervals. Costs are for one treatment.

Labour (10 person days @ $170) $1 700
Vehicle (3 days @ $36) $108
Consumables $200
Boat hire (3 days @$65) $195
Total $2 203

(b) Regeneration fire: May be required to stimulate seedling regeneration. Costs as for previous action, but without creation of fire break.

Labour (6 person days @ $170) $1 020
Vehicle (3 days @$36) $108
Consumables $200
Boat hire (3 days @$65) $195
Total $1 523

(c) Smoke treatment/water supplementation: May be required to stimulate and sustain seedling regeneration. Field apparatus for smoke treatment has been developed by Kings Park & Botanic Garden. Costs include apparatus, labour for planning, treatment and follow-up monitoring/watering and travel.

Labour (6 person days @ $170) $1 080
Labour (10 weeks @ $600) $6 000
On costs (28%) $1 680
Apparatus $1 000
Vehicle (11 weeks @$180) $1 980
Boat hire (3 days @$65) $195
Total $11 935

(d) Fencing: May be required to exclude macropods and prevent grazing of seedlings and resprouting plants. Costs include materials, transport and labour.

Planning & design (5days @$350) $1 750
Labour (10 days@$170) $1 700
Materials (500 m @$3.30) $1 650
Boat hire (8 days @$65) $520
Vehicle (2wks @$180) $360
Total $5 980

3.2 Disease avoidance and management.

ACTION 2.1 Undertake annual site inspections to detect invasion of P. cinnamomi.

Early detection of P. cinnamomi is essential if the impact of a disease epidemic on E. stuartii is to be mitigated. Annual monitoring is necessary to ensure that invasion of the pathogen and disease symptoms are detected early because death rates of Epacris species in infected areas may exceed 50% in the first year that symptoms become evident (Keith, unpubl. data).

Monitoring will entail a site inspection throughout the full extent of the population to record symptoms of disease. Soil samples will also be collected for analysis from any sites suspected to be infected and several representative apparently uninfected sites. The precise location of sample collection points will be recorded to assist with data interpretation. Monitoring will be carried out in January of each year because disease symptoms are most likely to be expressed by this time of year.

Disease monitoring will be integrated with population monitoring (Action 3.1.1), to be carried out at the same time of year. Integration of the two monitoring programs allows cost savings to be achieved. Costs of Action 3.2.1 are included within the estimates for Action 3.1.1.

ACTION 2.2 Promote and implement measures to ensure Phytophthora hygiene for all visitors to George III Historic Site and the northern section of Southport Lagoon Wildlife Sanctuary.

The major risk of disease introduction to George III Historic Site is through transport of infected mud on footwear of visitors including tourists, management staff and researchers. For this reason, visitation to the Historic Site will not be encouraged or promoted.

The Parks and Wildlife Service will ensure that all staff and external researchers follow hygienic procedures when visiting the site. These procedures include washdown or change of footwear at appropriate sites. The Recovery Team will distribute educational material to and seek the co-operation of local tourist ventures to encourage recreational bushwalkers to follow hygienic procedures when visiting the area. Key ventures to be targeted include the Ida Bay Railway Company and the Lune River Hostel. The educational material will include a brochure describing the impact and spread of the disease, describe hygienic washdown procedures and notify intending visitors that they will be required to change or washdown footwear at designated points. An existing brochure, 'Phytophthora Root Rot ... the plant killer' prepared by Parks and Wildlife Service with assistance from ANCA includes much of the necessary information. ESP funds will be required to prepare an insert with information locally applicable to the Southport area and George III Historic Site.

Appropriate signs designating washdown points will be erected at the Deep Hole Railway terminus, Southport boat ramp and at a location topographically suitable for washdown near the western boundary of George III Historic Site.

Parks and Wildlife Service and relevant research institutions will bear the costs of hygienic procedures for their own staff. Liaison with local tourist ventures will be carried out by local Service staff.

Action 2.2 costs
Year 1 2 3 4 5 6 7 8 9 10
Washdown equipment & chemicals $100 $100 $100 $100 $100 $100 $100 $100 $100 $100
Pamphlet printing $200                  
Liaison (2days@$350) $700 $700 $700 $700 $700 $700 $700 $700 $700 $700
Interpretation (1day@350) $350                  
Pamphlet insert (design & printing) $500                  
Production of 3 signs $700                  
Total $2 550 $800 $800 $800 $800 $800 $800 $800 $800 $800

ACTION 2.3 Development works to reduce the spread of infected mud.

Development of the access route to George III Historic Site will reduce the spread of infected mud and hence the risk of disease in the E. stuartii population. The current access track has developed in an unplanned manner determined by previous movements of vehicles and walkers. The aims of future access management will be to minimise the transport of mud to the Historic Site from elsewhere and to confine walker traffic within the Historic Site to a defined pathway. The preferred strategy is to redirect the existing access track, install a footwear washdown station and construct an elevated walkway within the Historic Site. Careful planning of these works is essential to ensure that the potential for invasion of P. cinnamomi is minimised.

The re-routed access track will direct walkers into an uninfected area which will serve as an 'clean' buffer zone through which all walkers are directed on their way to the Historic Site. A footwear washdown station with appropriate signage will be installed at the beginning of the new section of track to minimise the transport of infected mud into the buffer zone. The buffer zone should be as wide as possible, to minimise the transport of infected mud across its entire breadth, as a precaution against failure of walkers to effectively wash down their footwear. The new section of track required will therefore be up to 3 km in length and will be located to include a suitable washdown site and avoid areas likely to become muddy. The washdown station must be located at a topographically suitable site with permanently available water. A ground survey is necessary to locate sites of P. cinnamomi infection along the existing access route and to evaluate options for a new track route and a suitable site for the washdown station.

The purpose of the walkway is to confine walker movements to a single route within the Historic Site and to direct people away from significant stands of Epacris stuartii, thus minimising the chance of infection. The walkway will extend for up to 400 m from the western boundary of the Historic Site, across a drainage depression and across the headland summit to the George III Monument. With the co-operation of visitors, these works offer the best possible option for minimising the chance of introduction of P. cinnamomi into the Historic Site and, in the event that introduction occurs, containing the infection to a single route and hence minimising the chance of spread through E. stuartii habitat.

Planning and design of the works will be undertaken by Parks and Wildlife Service staff in consultation with the Recovery Team. Parks and Wildlife staff will carry out the ground survey and soil testing. Re-routing of the walking track and installation of the washdown station (a) will receive higher priority than construction of the elevated walkway (b).

Action 2.3 Year 1 costs
ExpenseCost
Planning & supervision (2 wks @$350) $1 050
Vehicle (1wk @$180) $540
Tools & equipment $2 200
Boast (6 days @ $65) $650
Track cutting (4 km @$1200) $4 800
Washdown station $1 000
Labour (27 days@$158 +15%) $4 907
Materials (400 m @$23) $9 400
Helicopter hire $4 200
Total $28 837

ACTION 2.4 Treatment and quarantine of infected areas

George III Historic Site will be quarantined at times considered appropriate by local management staff. Quarantine restrictions may be required when infections are evident in the vicinity of E. stuartii and when weather conditions pose a high risk of spread. If disease symptoms are reported within George III Historic Site and the presence of P. cinnamomi is confirmed (Action 3.2.1), infected vegetation and adjacent sites will be treated with phosphonate as soon as weather conditions are suitable. Initially, two treatments of phosphonate may be required, several weeks apart. Follow up treatments will be implemented at appropriate intervals. Parks and Wildlife Service has the capacity for rapid response to treat small areas to minimise the impact of disease.

Phosphonate is applied by spray to plant foliage and acts systemically to increase plant resistance to infection. It does not act directly on the pathogen. There is a possibility that phosphonate may elicit a toxic response when plants are exposed to multiple applications over time. There is no evidence of a toxic response in related Epacris species after one application (P. Barker, pers. comm. Forestry Tasmania), however there are no data on responses to multiple applications. It may therefore be necessary to monitor foliage of treated and control plants. The timing of, and necessity for phosphonate application will be determined by the appearance of the disease on site. Indicative costs are given for two treatments.

Action 2.4 costs
ExpenseCost
Labour (8 days@$350) $2 800
Equipment & chemicals $600
Vehicle (8 days@$36) $288
Monitoring (2days/year) $700
Foliage tests $300
Total $4 688

3.3 Research into regeneration ecology for population management.

ACTION 3.1 Determine seed longevity.

An understanding of the longevity of the soil seed bank is an essential basis for population management in the event of a major population decline. Existing germination data suggest that the population size could be restored through a regeneration fire to stimulate seed germination. However, the presence and activity of Phytophthora cinnamomi will determine the optimum timing of such a fire. If P. cinnamomi is active, new seedlings emerging after the fire will simply be exposed to the disease and die. The net result will be depletion of the soil seed bank. However, with time since the initial disease epidemic the activity of P. cinnamomi is known to decline. Thus seedlings may avoid disease if they emerge after the activity of the pathogen has declined. A delay of several years in implementing a regeneration fire may therefore be beneficial, so long as sufficient seeds remain viable in the soil over that time. Knowledge of seed longevity will establish the limits of flexibility to the timing of regeneration fires. There will also be spin off benefits for ex situ conservation. If seeds prove to be long-lived, seed storage would prove to be a cost-effective means of maintaining a very large genetically diverse ex situ collection.

A seed burial experiment will be carried out over two years to determine the longevity of seed in the soil. The Parks and Wildlife Service will provide supervision, vehicular transport and arrange use of laboratory facilities.

Action 3.1 Years 1 and 2 costs
Year 1 Expense Cost Year 2 Expense Cost
Supervision (5 days@$350) $1 750 4 days@$350 $1 400
Vehicle (3wks@$180) $540 2wks@$180 $360
Office equipt (10wks@$96) $960 6wks@$96 $576
Lab. space & equipt (52wks@$50) $2 600 52wks@$50 $2 600
Salary (10wks @ $775) $7 750 6wks@$775 $4 650
On costs (28%) $2 170   $1 302
Materials & Travel $300   $100
Total $16 024   $10 988

ACTION 3.2 Develop management technique for seedling regeneration.

Existing data indicate that seedling regeneration is limited by fire-related germination cues. Nonetheless, regeneration fires may fail to stimulate substantial seedling regeneration, as was the case after the 1994 fire. Laboratory experiments on E. stuartii seed (Section 1.5.5) and field trials on other species (Dixon et al. 1995) suggest that seedling regeneration may be enhanced in the field by application of smoke treatment followed by supplementary watering if soil moisture is limiting. If feasible, this may be an important management technique for restoring population size and one that is substantially more cost effective than reintroduction of ex situ stock. It is important that a regeneration technique is developed before the need arises to reverse a substantial decline in the wild population.

An experiment to test the effect of smoke treatment and supplementary watering on seedling establishment will be carried out in the field. The methodology will be developed in consultation with Western Australian researchers.

Year 1 costs
Expense Cost
Supervision (5 days@$350) $1 750
Vehicle (3wks@$180) $540
Office equipt (8wks@$96) $768
Salary (8wks @ $775) $6 200
On costs (28%) $1 736
Materials & Travel $1 000
Total $11 994

ACTION 3.3 Develop decision support system for management.

The are a potentially complex range of scenarios involving fire, disease epidemic and weather conditions to which management must respond to maintain Epacris stuartii above the population threshold. There are also a range of management responses available to restore population size, depending on the cause of decline (Action 3.2.1). A decision support system is required to assist wildlife and field staff to determine the most appropriate management response to various scenarios.

Year 1 costs
Expense Cost
Supervision (2 days@$350) $700
Office equipt (4wks@$96) $384
Salary (4wks @ $775) $3 100
On costs (28%) $868
Total $5 052

3.4 Ex situ safety net for wild population.

ACTION 4.1 (REVISED) Establish and maintain a representative ex situ collection of living plants at a suitable location remote from the wild population.

Under the worst case scenario, Epacris stuartii could decline to extinction in the wild during the term of this Plan. The development and maintenance of a genetically representative ex situ living collection is essential insurance against such an outcome. The principal requirements of such a collection are: (i) genetic representativeness; (ii) commitment to maintenance in perpetuity; and (iii) security against disease infection and vandalism. Genetic representativeness can only be assured by a genetic survey of the wild population. In the absence of such a survey, genetic representativeness may be approximated by stratified sampling of the population. Material for propagation should thus be collected from plants scattered throughout the entire population and from both microhabitat types.

The options proposed by Keith for ex situ conservation collection of Epacris stuartii are: (A) establishment and maintenance of a secure living collection at the Royal Tasmanian Botanical Gardens; (B) establishment of a living collection on other land; (C) establishment of a semi-natural population in suitable bushland; and (D) dispersal of propagated material through the community for cultivation in home gardens.

Option (A) was the preferred option for ex situ conservation but is the most expensive. A cheaper and possibly more cost effective alternative is a combination of options (C) and (D). Option (C), the establishment of an ex situ population on other land, was also proposed by Barker (1994) to be undertaken on Southport Island, adjacent to the wild population on Southport Bluff and consequently environmentally similar. Southport Island offers the security of an isolated population particularly with regard to Phytophthora cinnamomi infection.

Before proceeding with Southport Island as an ex situ site the island will be surveyed for botanical and zoological values. An assessment will be made with regard to the potential impacts of translocation of E. stuartii. The assessment will be undertaken with due regard to the translocation policy outlined in detail by ESAC. If other Epacris species are found on the island the potential impacts of hybridisation will also be closely studied. Details of the translocation will be lodged with the Threatened Species Unit at Environment Australia. Another location will be sought and assessed if Southport Island turns out to be unsuitable.

Since an ex situ wild population would not be closely monitored further security could be gained by also dispersing propagated material through the community for cultivation in home gardens (option D). These plants would be a source of material in the event of loss of the wild and ex situ populations.

Action 4.1 Years 2 and 3 costs
Expense Year 2 cost Year 3 cost
Salary, including oncosts and supervision $16 750 $15 750
Survey Southport Id. $1 000  
Propagation $6 000  
Establish wild population $22 000  
(labour accommodation and material)    
Distribution to private gardens and database $2 000  
Cryostorage and seed testing $6 000  
(Equipment, storage and technician)    
Office accommodation $1 200 $1 200
Vehicle and boat hire $1 625 $1 750
Total $26 575 $48 700

ACTION 4.2 Reintroduce a representative sample of plants to the wild at an appropriate time, should the wild population become extinct.

If Epacris stuartii becomes extinct in the wild, reintroduction will be necessary. To minimise site disturbance, unnecessary alteration to genetic composition of the population and costs, reintroduction from cultivated stock will not be considered until. Reintroduction will be attempted only when (i) extinction of the wild population is confirmed; (ii) attempts have been made to restore the wild population through management (e.g. stimulating regeneration from the soil seed bank); and (iii) when threatening processes that caused the decline of the wild population have been abated. A reintroduction project will follow guidelines endorsed by the Australian Network for Plant Conservation.

Extinction of the wild population is not expected during the term of this Plan. However, indicative costs are provided in the event that a funding application for reintroduction is required. A part-time project officer will be required for three years to plan the reintroduction project, co-ordinate planting and undertake monitoring and maintenance.

Action 4.2 Years 1,2 and 3 costs
Expense Year 1 Expense Years 2 & 3
Supervision (10 days@$350) $3 500 4 days@$350 $1 400
Vehicle (8wks@$180) $1 440 4wks@$180 $720
Office equipt (26wks@$96) $2 496 13wks@$96 $1 248
Propagation      
Salary (26wks @ $775) $20 150 13wks@$775 $10 075
On costs (28%) $5 642   $2 821
Materials $1 000    
Travel (20 trips@$25) $500 13 trips @$25 $325
Total $34 728   $16 579

Acknowledgements

Belinda Pellow, Wendy Potts, Tony Auld, Judy Scott and Brooke Craven assisted with fieldwork. Mark Fountain (Royal Tasmanian Botanical Gardens) assisted with collection of cuttings and successfully propagated the material. Stephen Harris provided advice and administrative support. Prof. Jamie Kirkpatrick made available laboratory space and equipment while I was a Research Associate of Department of Geography and Environmental Studies (University of Tasmania). Shauna Roche (Kings Park and Botanic Garden) supplied information, advice and materials for the germination experiment. Dr Phil Barker (Forestry Tasmania) made survey information available and participated in some fruitful discussion about the biology and management of Phytophthora in relation to Epacris. Tim Rudman (Parks and Wildlife Service) offered some very helpful ideas on management during and after a visit to the site. Transport to the site was assisted by the Lune River Youth Hostel and the Ida Bay Railway Company. This project was funded by the Australian Nature Conservation Agency under the Endangered Species Program (Project No. 423).


References

Barker, P. J. C. and Wardlaw, T. J. (1995). Susceptibility of selected Tasmanian rare plants to Phytophthora cinnamomi. Australian Journal of Botany 43, 379-386.

Bell, D. T., Rokich, D. P., McChesney, C. J. and Plummer, J. A. (1995). Effects of temperature, light and gibberellic acid on the germination of seeds of 43 species native to Western Australia. Journal of Vegetation Science 6, 797-807.

Bradley, P. (1994). 'Fire report - Southport Lagoon, 2 April 1994.' Unpublished report. Tasmanian Parks and Wildlife Service.

Briggs, J. D. and Leigh, J. H. (188). 'Rare or threatened Australian plants.' Special Publication No. 14. Australian National Parks and Wildlife Service, Canberra.

Crowden, R. K. and Menadue, Y. (1990). Morphometric analysis of variation in the 'Epacris tasmanica complex' (Epacridaceae). Australian Systematic Botany 3, 253-264.

Curtis, W. M. (1963). 'Flora of Tasmania. Part 2.' Tasmanian Government Printer.

Dixon, K., Roche, S. and Pate, J. S. (1995). The promotive effect of smoke derived from burnt native vegetation on seed germination of Western Australian plants. Oecologia 101, 185-192.

Keith, D. A. (1991). 'Coexistence and species diversity in upland swamp vegetation: the roles of environmental gradients and recurring fires.' PhD thesis. University of Sydney.

Keith, D. A. (1995). Surviving fire and fungal pathogens: are there life-history solutions for threatened species of Epacris? In 'Bushfire '95: Proceedings of the fifth Australian Bushfire Conference' (Eds. T. Blanks, M. Chladil, R. Chuter and K. Green). (Forestry Tasmania, Parks and Wildlife Service, Tasmania Fire Service: Hobart)

Keith, D. A. (1996). Fire-driven extinction of plant populations: a review of theory and evidence from Australian vegetation. Proceedings of the Linnean Society of New South Wales 115, in press.

Leigh, J. H., Briggs, J. D. and Hartley, W. M. (1981). 'Rare or threatened Australian plants.' Special Publication No. 7. Australian National Parks and Wildlife Service, Canberra.

Skinner, A. D. (1981). 'Southport Lagoon Conservation Area - Fire February 1981.' Unpublished report. Tasmanian Parks and Wildlife Service.


APPENDIX: Distribution of Epacris stuartii on Southport Bluff.

NOT FOR DISTRIBUTION

For permission to use, copy or distribute this map, contact Senior Botanist, Parks and Wildlife Service, GPO Box 44A, Hobart 7001 ph. (03)62336556.

Epacris stuartii is essentially confined to a narrow strip around the rim of the headland: up to a few metres behind the cliff tops; in crevices on the cliff faces; up to a few metres from the cliff bases; and in small gullies that interrupt the cliff line. Except for two locations where bedrock has been exposed some 30-40 m from the rim, E. stuartii is not found on the inclined headland summit more than a few metres from the rim. Plants are scattered mostly around eastern and southern aspects in numerous small subcatchments draining different parts of the headland.

The preparation of this Recovery Plan was funded by the Australian Nature Conservation Agency under the Endangered Species Program (Project No. 423).

Copyright © 1996 Australian Nature Conservation Agency, GPO Box 636, Canberra 2601, A.C.T.

Cover Photograph: Epacris stuartii in flower September 1995, by Dr D. A. Keith

Published by: Tasmanian Parks and Wildlife Service, GPO Box 44A, Hobart 7001, Tasmania.