Vegetation types described for Bega Valley Shire - NSW Government

State of the Vegetation Report 

for the 

Bega Valley Shire 

Prepared for Bega Valley Shire Council and the 

Far South Coast Catchment Management Committee 

Prepared by Jackie Miles 

October 2000

State of the Vegetation Report 

for the Bega Valley Shire 

Prepared for Bega Valley Shire Council and the 

Far South Coast Catchment Management Committee 

by Jackie Miles 

November 2000

© Bega Valley Shire Council 2000 

This work is copyright. Apart from any use as permitted under the Copyright Act 1968, no part may be 

reproduced by any process without prior written permission from the publisher. 

Published by: 

Bega Valley Shire Council 

PO Box 492 

Bega NSW 2550 

Phone: 02 6244 2222 

Fax: 02 6499 2200 

Email: council@begavalley.nsw.gov.au 

Edited and designed by Green Words & Images, Canberra 

Printed by Panther Publishing and Printing, Canberra

State of the Vegetation Report for Bega Valley Shire 

Table of contents 

Acknowledgements 6 

Note 6 

Executive summary 7 

Assessment of available vegetation data 7 

State of the vegetation report 9 

Conservation status of vegetation types 9 

Threats to vegetation 12 

Vehicles for vegetation protection 14 

Threatened flora in Bega Valley Shire 15 

Recommendations 15 

Part A: Assessment of data currently available on vegetation in Bega Valley Shire 

1 Introduction 21 

1.1 Comprehensive regional assessment data 21 

1.2 Non-CRA sources of information on vegetation 22 

2 Assessment of accuracy and usefulness of CRA-modelled vegetation 

distribution map 24 

2.1 Derivation of the vegetation types 24 

2.2 The vegetation types 25 

2.3 Description of the vegetation types with indicator species tables 26 

2.4 Ease of recognition in the field of the vegetation types 29 

2.5 Vegetation types in which the range of species assemblages is inadequately described 31 

2.6 Derivation of the CRA-modelled vegetation maps 31 

2.7 Checks on map accuracy during production 33 

2.8 Assessment of map accuracy at fine scale 33 

3 Comparison of aerial photo interpretation vegetation typing with CRA vegetation model 39 

3.1 Derivation of aerial photo interpretation map 39 

3.2 Comparison of API and CRA vegetation map for specific locations 40 

4 Conclusions 42 

Part B: State of the vegetation report 

1 Introduction 44 

2 Conservation status of the vegetation types 47 

2.1 Criteria for adequacy of conservation status 47 

2.2 Conservation status of vegetation types in Bega Valley Shire 48 

3 Potential threats to vegetation and their impacts 56 

3.1 Agricultural landscapes 56 

3.1.1 Fragmentation and clearing 56 

3


4 

3.1.2 Dieback 57 

3.1.3 Loss of biodiversity – flora 58 

3.1.4 Loss of biodiversity – fauna 61 

3.1.5 Weed invasion 62 

3.1.6 Impact of nutrient-laden run-off and sedimentation on wetland vegetation 64 

3.2 Currently forested landscapes 64 

3.2.1 Clearing 65 

3.2.2 Timber harvesting 66 

3.2.3 Fire regimes 68 

3.2.4 Dieback 70 

3.2.5 Weed invasion 71 

3.2.6 Phytophthora 71 

4 Protection mechanisms for vegetation 73 

4.1 Native Vegetation Conservation Act 1997 73 

4.2 Threatened Species Conservation Act 1995 73 

4.3 Fisheries Management Act 1994 74 

4.4 State Environmental Planning Policy 14 75 

4.5 State Environmental Planning Policy 44 75 

4.6 Department of Land and Water Conservation dam licensing requirements 75 

4.7 Bega Valley Local Environmental Plan 76 

4.8 Extension programs and incentive schemes 76 

5 Threatened flora in Bega Valley Shire 78 

5.1 Species formally listed as rare or threatened 78 

5.2 Regionally uncommon species 79 

6 Conclusions 81 

7 Recommendations 83 

7.1 For Bega Valley Shire Council 83 

7.2 For the Catchment Management Committee 86 

7.3 For the Regional Native Vegetation Committee 87 

List of plant common names 88 

List of acronyms 91 

Glossary 92 

Bibliography and references 94 

Appendices 

1 Vegetation types described for Bega Valley Shire 100 

2 Detailed assessment of map accuracy for selected vegetation types 111 

3 Vegetation types which can be split into further sub-types 122 

4 Environmental weeds of Bega Valley Shire 125


5 Threatened and rare plants of Bega Valley Shire 133 

6 Species of regional conservation significance 136 

7 Remnant vegetation in Bega Valley Shire on roadsides and other public lands administered 

by Council 141 

Map 1: Study area with land tenure as at 1 January 1999 20 

Map 2: Depleted vegetation types on private land 50 

Table 1: Priority for protection of vegetation types (from Eden Regional Forest Agreement) 10 

Table 2: Suggested amended priorities for protection of vegetation types 11 

Table 3: Sample diagnostic species description 27 

Table 4: Estimated area of vegetation types and proportion in reserves 35 

Table 5: Priority for protection of vegetation types 52 

Table 6: Environmental weeds of Bega Valley Shire 125 

Table 7: Threatened and rare plants of Bega Valley Shire 133 

Table 8: Species of rain shadow valley dry grass forests 136 

Table 9: Riparian or wetland species 137 

Table 10: Species known mostly from private property around rain shadow valley margins 139 

Table 11: Species occurring mostly on the coast and known only from privately owned land 

or vacant crown land or with a significant proportion of records from private property 

with some occurrences in reserves as well 140 

5


Acknowledgements 

First and foremost, thanks go to David Keith of the 

New South Wales National Parks and Wildlife 

Service (NPWS) for maintaining his patience and 

good humour in the face of interminable questions 

and criticisms of the vegetation model. Also to 

Graham Roche at Bega Valley Shire Council (BVSC) 

for production of maps and advice on ArcView, and 

to Don McPhee at the Department of Land and 

Water Conservation (DLWC) without whose initial 

enthusiasm the funding for this project would not 

have been obtained. 

Thanks also to the numerous people who provided 

information for the report, particularly Steve Dovey 

Note 

Since this report was completed, the Far South Coast 

Catchment Management Committee (FSC CMC) 

has been replaced by the South East Catchment 

Management Board, covering the wider area of the 

Far South Coast (Bega Valley Shire), Lower South 

Coast (Eurobodalla Shire) and Snowy-Genoa CMC 

6 

from NPWS, who helped with the section on fire 

regimes, Alan Smith (BVSC) for advice and 

feedback about weeds and Max Beukers (NPWS) for 

information about the Comprehensive Regional 

Assessment (CRA) methodology. 

Comment on an early draft was provided by David 

Keith, Justin Gouvernet (DLWC), Vic Jurskis (State 

Forests of New South Wales) and Heather Stone. 

Finally, thanks to the many landholders who allowed 

access to their properties for details of vegetation 

distribution to be checked. 

regions. References to past activities of the 

FSC CMC still apply to that body, but 

recommendations for future activities should now be 

taken to be addressed to the South East Catchment 

Management Board.


Executive summary 

This report has the following aims: 

• to interpret for a lay audience the information 

about vegetation which has become publicly 

available as a result of the recent Comprehensive 

Regional Assessment (CRA) in the Eden region; 

• to assess the accuracy and usefulness of the CRA 

vegetation maps, principally as they relate to 

vegetation types found primarily on private 

property; 

• to provide an overview of other information 

relating to vegetation in Bega Valley Shire and 

the area covered by the Far South Coast 

Catchment Management Committee; 

• to provide a comprehensive State of the 

Vegetation Report covering: 

– the conservation status of the various 

vegetation types described for Bega Valley 

Shire; 

– threats to vegetation; 

– information about threatened plant species 

in the Shire; and 

• to make recommendations regarding future 

vegetation management directions in the Shire. 

Assessment of available 

vegetation data 

Prior to the CRA there was very little information 

readily available on the vegetation of Bega Valley 

Shire. State Forests of New South Wales (SFNSW) 

had developed a State-wide vegetation classification 

scheme and applied it to the mapping of some State 

Forests in the region. The National Parks and 

Wildlife Service (NPWS) had undertaken vegetation 

sampling throughout the region which was 

ultimately to produce the CRA vegetation model. A 

number of papers and reports describe earlier phases 

of this work but do not include vegetation maps 

(that is, publicly available maps). As a result of the 

CRA process there are now two alternative 

vegetation maps available for the Shire in digital 

format on ArcView. They are held locally by 

Merimbula NPWS, SFNSW, Bega Valley Shire 

Council and the South East Forests Conservation 

Council. A hard-copy map accompanies the paper 

by Keith and Bedward (1999). 

The product on which this report concentrates, the 

CRA pre-1750 vegetation map and extant vegetation 

map, is derived from about 1500 samples 

(20 x 20 metre plots) of vegetation from throughout 

the Eden Management Area (see Map 1, p. 20, for 

location of the EMA). The method records all plant 

species present in sample plots and their relative 

abundance. It uses PATN cluster analysis (Belbin 

1994) to derive vegetation types from the samples, 

by clustering samples according to their degree of 

similarity. The analysis resulted in the description of 

79 vegetation types for the Eden Management Area 

of which only nine occur wholly outside the Bega 

Valley Shire boundary. The Eden Management Area 

includes an area of the Southern Tablelands, west to 

Numeralla and Bombala, but that area is not covered 

in this report. 

Data recorded about site conditions on the sample 

plots were used to create a set of decision rules 

describing where in the landscape the vegetation 

types could be expected to occur. The rules used 

features such as geology, altitude, temperature and 

rainfall variables and local topographic position. 

Various geographic information system (GIS) layers 

(geology, topography, modelled climate) were used 

to build up a model of vegetation habitat in the 

region. Air photo interpretation (API) was used to 

map non-eucalypt-dominated vegetation types (for 

example, heaths, rainforest, riparian scrubs, estuarine 

communities). The application of the decision rules 

to the GIS habitat data layers produced a model of 

where the described vegetation types are likely to 

occur in the district (the pre-1750 map). 

Superimposing a GIS vegetation layer (the ‘woody/ 

non-woody’ layer), derived from 1994 Landsat 

imagery, excised areas which had been cleared or 

carried plantations, to produce the extant vegetation 

map. The production of the vegetation model and 

the vegetation types are described in Keith and 

Bedward (1998, 1999). 

7


The second vegetation mapping product available 

for the region is a broad forest classes map derived 

solely from API. Vegetation types used here are based 

on tree species only rather than full floristics, and are 

derived from Research Note 17 (FCNSW 1989). 

They are generally not directly equivalent to the 

vegetation types derived from full floristics by Keith 

and Bedward. A number of air photo interpreters 

mapped the location of these vegetation types using 

a limited amount of field work to assess whether 

their interpretation of the vegetation types from the 

air photos had been correct. Because of time 

constraints on this project not all of the district was 

mapped to the same level of detail. 

The two vegetation products each have their 

advantages and disadvantages. In the field it may be 

easier to allocate the vegetation to a broad forest class 

than to one of the Keith and Bedward vegetation 

types, because the former are often broader 

categories and are based only on tree species. To use 

the Keith and Bedward classification requires a high 

level of local botanical knowledge. The criticism has 

been levelled at the Keith and Bedward classification 

that vegetation types have been too finely divided 

and that some are indistinguishable in the field. 

However, in attempting to use the classification it 

was found that some vegetation types could actually 

be more finely divided into recognisable sub-types. It 

may therefore be largely a matter of familiarity with 

the system as to how well its description of regional 

vegetation is perceived. However, given the inherent 

variability of vegetation associations in the field, in 

response to site factors such as topography and soil 

conditions, there will always be some areas of 

vegetation which are difficult to assign to one of the 

described types. 

The greatest difference between the modelled and 

API-derived vegetation maps is that the 

accompanying documentation for the model (Keith 

& Bedward 1998, 1999) provides information about 

the degree to which each vegetation type is estimated 

to have been cleared, and the proportion predicted 

to occur in reserves, State Forests and on private 

property. This information is crucial in guiding land 

use decisions in the region but is not provided with 

the API map. Because of the non-correspondence of 

many of the broad forest classes with the Keith and 

Bedward vegetation types, it is difficult to use the 

8 

API map in conjunction with the information about 

conservation significance provided by Keith and 

Bedward. Because conservation significance is 

generally the factor of greatest interest, it may not be 

important that it is sometimes impossible to assign 

vegetation to a particular Keith and Bedward type in 

the field. Vegetation types that are hardest to tell 

apart are likely to be of similar conservation 

significance. 

The overall accuracy of the API map has not been 

assessed at regional or local (within one kilometre 

radius) scale. The CRA model has been assessed at 

this scale during production, by withholding 10 per 

cent of the vegetation samples from production of 

the model and using them to check the accuracy of 

prediction for particular vegetation types. It was 

found to be reasonably accurate at these scales. The 

fine-scale accuracy of the CRA model has been 

checked for this report in order to determine how 

useful it might be for assisting in land use decisions 

at the single property scale. Most checking was done 

of vegetation types which occur predominantly on 

private property, since these are the most depleted 

and least reserved vegetation types in the region. The 

accuracy of the model and the API map were also 

compared for a single 120 hectare area at Brogo. 

A number of inaccuracies were found with the CRA 

model at fine scale. Generally, depleted vegetation 

types that formerly occurred over large expanses in 

agricultural areas were found to be even more 

depleted than the map indicated. Many small 

occurrences mapped in reserves, generally around 

the margins of the agricultural areas, were found to 

be incorrect (that is, these types had been 

over-mapped compared with their actual 

occurrence). For some vegetation types which are 

naturally more rare and fragmented, such as those 

confined to drainage lines, the map was found to 

underestimate the original extent, largely due to 

failure to detect the vegetation types from API. 

Feedback from this project has resulted in the 

addition of stands of the vegetation types Dry 

Rainforest, Riverine Forest and Floodplain Wetland 

on private property to Version 4 of the map (Keith 

& Bedward 1999). This version is now available to 

Council in digital format. However, this does not 

mean that all occurrences of these vegetation types 

on private property are now recorded.


Comparison of the map derived from the vegetation 

model and the API-derived map in a particular area 

where the vegetation types and their distribution 

were well known showed that neither was 

particularly accurate. Each missed detecting small 

areas of restricted vegetation types, such as Dry and 

Warm Temperate Rainforests and Riverine Forest, 

though this is understandable, given the small size of 

the stands present. Each also missed detecting a 

vegetation type present over a large part of the 

property. The API map did not detect any forest red 

gum-dominated vegetation types (Brogo Wet Vine 

Forest or Bega Dry Grass Forest) and the model 

missed Escarpment Dry Grass Forest. Neither map 

had the vegetation types accurately located within 

the property, although both were correct in some 

parts. 

It therefore seems that at fine scale neither map can 

be relied on to correctly predict the vegetation types 

present, or their distributions. The Keith and 

Bedward model describes vegetation types more 

thoroughly than does the broad forest classes 

classification, which uses only trees to define its 

forest classes. Only the Keith and Bedward model 

provides information about the conservation 

significance of the vegetation types. Given its 

reasonable level of accuracy at regional and local 

scale, it can therefore be used to give an indication of 

broad areas in which vegetation types of greatest 

conservation significance are likely to occur. This 

will provide useful information for future vegetation 

management for Council, the Catchment 

Management Committee, the Department of Land 

and Water Conservation and the Regional Native 

Vegetation Management Committee (when 

appointed). However, for planning decisions at the 

individual property scale, field survey will still be 

required to determine what vegetation type is 

present on a particular site. The Keith and Bedward 

classification also provides a framework in which to 

place observations of plant species. Site assessment 

may now provide the name of the vegetation type 

present on a site, and some information about the 

site’s conservation significance, rather than a simple 

list of the plant species present. 

Brief descriptions of the 70 relevant vegetation types 

described by Keith and Bedward can be found in 

Appendix 1 (p. 100) of this document. Appendix 2 

(p. 111) gives details of fine-scale map accuracy for 

selected vegetation types and Appendix 3 (p. 122) 

describes vegetation types which may not have been 

adequately captured by the sampling process, and 

hence could be further subdivided into more precise 

vegetation types. 

State of the vegetation report 

Conservation status of vegetation types 

The vegetation types found in Bega Valley Shire and 

their estimated degree of clearing and reservation are 

listed in Table 4 (p. 35) of this document. However, 

figures for clearing and the proportion of vegetation 

types found on different land tenures may not be 

very accurate, as they are based on modelled 

vegetation distribution, which was shown to contain 

numerous minor errors at the fine scale. Each 

vegetation type has been allocated a number and a 

name which describes its structural features (heath, 

scrub, forest), the nature of the understorey (grassy 

or shrubby) and features of the habitat or a 

particular location where it occurs. Keith and 

Bedward (1998) group the vegetation types 

according to their degree of floristic similarity, but 

this report groups them more around structural 

features, into the following categories: 

• rainforests; 

• scrubs, dominated by shrubs or small trees, 

occurring on rocky or other dry infertile sites; 

• heaths, dominated by sclerophyll shrubs, 

sometimes with sedges abundant in wet sites; 

• estuarine communities (seagrasses, mangroves, 

salt marsh and so on); 

• riparian and wetland communities; 

• wetter eucalypt forests of the escarpment and the 

more sheltered coastal sites; 

• dry grass forests found in the dry rain shadow 

valleys of the agricultural areas and on granitoid 

substrates in the hinterland; and 

• dry shrub and intermediate shrub forests found 

on the more exposed sites and poorer soils at all 

elevations throughout the region. 

9


Most of the dry and intermediate shrub forests and 

the wetter eucalypt forests are adequately reserved in 

the region, in that they meet the agreed criterion of 

having 15 per cent of their estimated original extent 

in conservation reserves (JANIS 1996). Those 

vegetation types which are naturally rare or 

fragmented, such as rainforests and some riparian 

types, have been allocated a reservation target of 

60 per cent or 100 per cent of existing stands in 

reserves (JANIS 1996). Generally this cannot be met 

by transferring production forest into reserves 

because of the scattered distribution of these 

vegetation types. These reservation targets are 

therefore being met under the Eden Region Forest 

Agreement by creation of an informal reserve 

network from which logging will be excluded within 

State Forests, or by management by prescription in 

the case of rainforests. 

The vegetation types which are most depleted and 

are least well conserved in the region are, not 

surprisingly, those associated with the agricultural 

areas. These are four formerly widespread forest 

types: 

• Bega and Candelo Dry Grass Forests 

(20 and 21); 

• Bega Wet Shrub Forest (19); and 

• Brogo Wet Vine Forest (18). 

10 

There are several vegetation types associated with 

drainage lines: 

• Northern Riparian Scrub (39); 

• Riverine Forest (40); and 

• Floodplain Wetland (60). 

And there is the naturally rare and fragmented Dry 

Rainforest (1) which is restricted to rocky sites on 

upper slopes and gully heads. Of these depleted 

vegetation types, 18, 20, 21 and 40 are found 

exclusively in agricultural areas in this region. 

Flats Wet Herb Forest (17) is mapped as having 

about half of its naturally uncommon and 

fragmented occurrence on private property in the 

Wyndham area (the accuracy of this was not 

checked), as well as occurrences in State Forests and 

reserves. It generally occurs on flats associated with 

drainage lines at higher elevations. 

Dune Dry Shrub Forest (36) occurs on sandy soils 

very close to the sea. This habitat is scarce in Bega 

Valley Shire, with much of the coastline being rocky, 

so the vegetation type is naturally rare in the region. 

Its extent was found to have been overestimated by 

the model, as was its degree of reservation. It occurs 

on vacant crown land and private property 

immediately south of Bermagui, and on private 

property scattered along the coast, with only a few 

very small occurrences in reserves. Some of the 

crown land occurrences have subsequently been 

transferred into reserves. 

Table 1: Priority for protection of vegetation types (from Eden Region Forest Agreement) 

High priority Moderate priority Low priority 

19, Bega Wet Shrub Forest 1, Dry Rainforest 22A, Monaro Dry Grass Forest 

20, Bega Dry Grass Forest 17, Flats Wet Herb Forest 30, Wallagaraugh Dry Grass Forest 

21, Candelo Dry Grass Forest 18, Brogo Wet Vine Forest 

40, Riverine Forest 36, Dune Dry Shrub Forest 

60, Floodplain Wetland 39, Northern Riparian Scrub 

23A, Monaro Grassland 23B, Monaro Basalt Grass 

Woodland 

24, Subalpine Dry Shrub Forest 22B, Numeralla Dry Shrub 

Woodland


Table 2: Suggested amended priorities for protection of vegetation types 

High priority Moderate priority 

60, Floodplain Wetland 19, Bega Wet Shrub Forest 

21, Candelo Dry Grass Forest 63, Estuarine Wetland Scrub 

20, Bega Dry Grass Forest 17, Flats Wet Herb Forest 

40, Riverine Forest 6, Coastal Warm Temperate Rainforest 

39, Northern Riparian Scrub 7, Hinterland Warm Temperate Rainforest 

18, Brogo Wet Vine Forest 64, Salt Marsh 

1, Dry Rainforest 65, River Mangrove 

36, Dune Dry Shrub Forest 66, Grey Mangrove 

Estuarine vegetation types are also thought to have 

been depleted by clearing for agriculture and 

residential development, and are not well reserved. 

However, the Fisheries Management Act 1994 and 

State Environmental Planning Policy 14 do provide 

some measure of protection for these vegetation 

types outside reserves (see Part B, section 4). 

The Eden Region Forest Agreement document 

(Anon 1999, Table 1, page 15) lists priorities for 

conservation in the region which cannot be achieved 

on public lands. These are reproduced in Table 1. 

Of the listed vegetation types shown in Table 1, the 

last two types in each column (including both low 

priority types) have all or nearly all of their 

occurrences outside Bega Valley Shire and are hence 

outside the scope of this report. Due to the errors 

which were found in mapping of the other 

vegetation types that do occur within Bega Valley 

Shire during field work for this report, some 

alterations in priority are suggested. With the 

exception of Flats Wet Herb Forest (17), which was 

not checked, nearly all of the listed vegetation types 

were found to have been over-predicted compared 

with their actual distribution. That is, they are even 

more depleted and less well reserved than the CRA 

information suggests. 

Native grasslands are unlikely to have occurred 

naturally in the Shire, but there may be areas of 

secondary grassland produced by clearing of grassy 

forests which contain regionally uncommon species 

of herbs and grasses. The CRA vegetation mapping 

does not indicate the location of such areas, since its 

mapping of remnant vegetation in agricultural areas 

is based on the presence of reasonably dense stands 

of trees. These areas are likely to represent former 

stands of Bega or Candelo Dry Grass Forest. 

Table 2 lists this report’s suggested priorities for 

conservation of remnant vegetation on private 

property in Bega Valley Shire. 

It should be remembered however, that membership 

of one of the above vegetation types should not be 

the sole criterion for targeting an area of remnant 

vegetation for conservation. Remnant vegetation 

may belong to a type which is of high conservation 

significance, yet still be so severely degraded that 

resources would be better directed to another site. 

Factors which need to be taken into account in 

assessing the conservation significance of remnant 

vegetation are: 

1. Vegetation type. The most significant types are 

listed in Table 2 above. 

2. Diversity of native species. Given that the 

grassy forest types carry a high proportion of 

their species diversity in the ground cover layer, 

this can really only be determined by someone 

with a thorough knowledge of local flora. 

Wetlands present a similar degree of difficulty in 

distinguishing native from introduced flora. 

Generally it is safe to assume that the higher the 

native species diversity on a site, the less 

degraded is the remnant. However, some plant 

communities (for example, Phragmites reed beds) 

have a naturally low level of species diversity, but 

are still of conservation significance. 

11


3. Degree of weed invasion. Generally the lower 

the proportion of introduced species, the higher 

will be the conservation significance of the 

remnant. However, some level of introduced 

flora presence is inevitable in remnant 

vegetation, and a higher weed level could be less 

important than say, the presence of significant 

species. 

4. Presence of regionally significant species. The 

presence of any of the species listed in 

Appendix 6 (p. 136) would increase the 

significance of any remnant stand. The greater 

the number of such species present, the greater 

the conservation significance. Plant species 

actually listed as rare or threatened in 

Commonwealth or State threatened species 

legislation are unlikely to occur on private 

property in Bega Valley Shire, although the 

possibility should not be discounted altogether. 

5. Structural diversity. Stands of greater structural 

diversity (for example, with several vegetation 

layers present, more than one age class of tree, 

litter and woody debris present) will generally be 

of more value to fauna and are likely to be in 

better health than stands with a simpler 

structure. However stands with only one or two 

layers, such as Phragmites reed beds, secondary 

grassland or eucalypt woodland with a grassy 

understorey, may still be of high conservation 

significance. 

6. Age structure of the stand. The presence of 

more than one age class of tree is desirable. Old 

trees and dead stags provide habitat for 

hollow-dependent fauna. Mature trees provide 

for continuity of the supply of hollows as well as 

foraging and nesting habitat. Regenerating trees 

provide for the long-term persistence of the 

stand. Stands consisting only of senescent trees 

(which may be beyond reproducing) or only of 

regeneration may be of lower conservation 

significance. However, stands containing only 

mature trees would not be of lower significance, 

since site conditions could be managed to 

produce regeneration. 

7. Isolation or contiguity of the stand. In an area 

where remnant vegetation is abundant, one 

particular stand may be regarded as expendable 

12 

depending on how it scores on the other factors 

discussed here. Conversely, in an area where 

remnant vegetation is sparse or degraded, each 

stand is important, so that stands of lower 

quality might be accorded a higher conservation 

significance than they would elsewhere. A stand 

which is strategically located so as to be a 

potential wildlife corridor will be of greater value 

than an isolated stand, all other things being 

equal. Stands which are connected to other 

vegetation are also more likely to remain viable 

in the long term. 

8. Size of stand. In general, the larger the better. 

Very small stands are unlikely to sustain the 

ecosystem processes necessary to keep them 

viable. However, given a reasonable level of 

species diversity, small stands can be gradually 

enlarged by encouraging natural regeneration or 

by buffer plantings of appropriate species. 

Threats to vegetation 

In agricultural areas most of the clearing was 

completed many years ago and the trend is now 

towards regeneration on marginal agricultural lands 

around the valley margins. On ungrazed or lightly 

stocked areas, such as roadsides and steep banks, 

regeneration is occurring throughout agricultural 

areas, as long as relict trees remain to provide a seed 

source. 

Threats to the persistence of remnant vegetation in 

agricultural areas include loss of understorey, lack of 

regeneration, dieback and weed invasion. 

Uncontrolled livestock access to remnants 

contributes to these problems through the trampling 

and browsing of native vegetation, nutrient 

enrichment and compaction of soils, disturbance to 

run-off patterns and erosion. Fencing of significant 

remnant vegetation needs to be given a high priority. 

Some livestock access may be desirable as a 

vegetation management tool, but the timing and 

intensity needs to be controllable. The use of fire in 

agricultural landscapes also needs consideration, 

since fire timing and frequency may have important 

impacts on vegetation. Most remnant vegetation in 

agricultural areas would be burnt very infrequently. 

As a result of the clearing of native vegetation for 

agriculture and the degradation of remnant stands,


there have been substantial losses of biodiversity in 

the Shire, though these have not been as severe as in 

inland areas. The extinction and near-extinction of 

some mammal species following European 

settlement of the region has been documented 

(Lunney & Leary 1988) and many other species of 

both fauna and flora have no doubt declined. 

However, there are no species of fauna listed as 

threatened under the New South Wales Threatened 

Species Conservation Act 1995 which are directly 

dependent on habitats provided within agricultural 

areas in the Shire. This stands in contrast to the 

Southern Tablelands and Western Slopes where 

species such as eastern earless dragon, striped legless 

lizard, regent honeyeater and superb parrot depend 

on remnant native grasslands or woodlands which 

are almost entirely unreserved. With the exception of 

three species of shrub (Zieria formosa, Z. buxijugum 

and Z. parrisiae) found in very restricted areas at 

Lochiel, there are no plant species listed in the 

Threatened Species Conservation Act 1995 which are 

likely to occur on private property in Bega Valley 

Shire. However, there are numerous regionally 

uncommon plants which are largely confined to 

remnant vegetation in agricultural areas (see 

Appendix 6, p. 136). 

In areas which are still forested, possible threats on 

private property include: 

• clearing for residential development or 

agriculture; 

• inappropriate fire regimes; 

• invasion by environmental weeds adjacent to 

residential areas; and 

• dieback caused by bell miner colonies. 

On the coastal strip most of the widely distributed 

vegetation types are adequately reserved, although 

more restricted types such as rainforest or Floodplain 

Wetland (60) may be affected by the threats listed 

above. Clearing on the coast is more likely to be 

significant for its impact on fauna populations, since 

the forests of the region support numerous 

forest-dependent species which are listed as 

threatened in the Threatened Species Conservation Act 

1995. While abundant forest remains on public 

lands in the Shire, its capacity to support 

forest-dependent fauna may have been reduced by 

timber harvesting activities, both in State Forests and 

in those National Parks which were logged before 

their change of tenure. Many National Parks also 

contain areas which were formerly cleared for 

agriculture, or selectively logged, so that the forest in 

these areas consists primarily of young trees. This is 

particularly significant for fauna which require tree 

hollows for shelter and breeding sites. 

Forests on private land may therefore have a role to 

play in fauna conservation in the region. In some 

instances such forests may actually provide better 

fauna habitat than nearby production forests or 

previously disturbed reserves. Privately owned forests 

may also be strategically located so as to provide 

links between public forests. These links may enable 

the movement of fauna species which are not able to 

cross cleared or lightly wooded areas between areas 

of potential habitat. This movement can be 

important for genetic exchange between 

populations, and to enable recolonisation of an area 

from which a population has been eliminated by, for 

example, wildfire. There are numerous areas in the 

Shire which have this potential to function as 

wildlife corridors. They are listed in Part B, 

section 7.1. Vacant crown land and public forests 

contribute significantly to some of these corridor 

areas, as well as private property. 

In addition to potential impacts on fauna 

populations, timber harvesting in both private and 

public forests may have some direct impacts on 

species composition of forests. Studies in dry 

silvertop–stringybark forest types in the Eden 

Management Area and East Gippsland indicate that 

these forests are relatively resilient to disturbance 

(Bridges 1983; de Chazal 1992; Mueck & Peacock 

1992; Loyn et al. 1983). Species composition in 

these forests appears largely unchanged by logging. 

However, studies in wetter forest types suggest that 

these forests could show some long-term changes in 

composition due to logging (Mueck & Peacock 

1992; Ough & Ross 1992). Research by SFNSW in 

the Eden Management Area has been restricted to 

silvertop–stringybark forest types. The long-term 

impacts of frequent fuel reduction fires, a practice 

associated with integrated logging, are unknown at 

this stage. SFNSW has been conducting research, 

again only in silvertop–stringybark forest types, but 

the data has not yet been analysed. 

13


Bell miners, by causing dieback in trees within and 

around the margins of forests in the region, may 

have an appreciable impact on vegetation. 

Observation on the northern and central New South 

Wales coast has shown an apparent link between 

forest disturbance and bell miner colony location 

and persistence (C Stone, SFNSW, pers. comm.). 

Relatively rapid change in tree species composition 

has also been observed in those areas. There is an 

urgent need for the collection of baseline data on 

bell miner location and impacts across all land 

tenures in this region. In the interim there may be 

some justification for culling bell miners in areas 

where they can be seen to be affecting vegetation of 

high conservation significance. 

Vehicles for vegetation protection 

Legislation intended to provide protection for 

vegetation in New South Wales is reviewed in Part B, 

section 4. In general it seems that such legislation 

currently provides little protection for the most 

depleted vegetation types in Bega Valley Shire. The 

Native Vegetation Conservation Act 1997 introduced a 

requirement to obtain consent from the Department 

of Land and Water Conservation for clearing of 

native vegetation, except in accordance with a 

Regional Vegetation Management Plan. Such a plan 

does not yet exist for this region. However, because 

of the exemptions which apply, the Act actually 

provides little protection to the vegetation types 

which most need it – that is, those which are 

reduced to small fragmented remnants. 

The Fisheries Management Act 1994 provides for 

protection of mangroves and seagrasses, but not for 

other estuarine vegetation such as salt marsh and 

estuarine melaleuca scrub. Many estuarine areas are 

declared protected wetlands under State 

Environmental Planning Policy 14, but this 

instrument does not provide any protection for their 

catchments, only for the wetland itself as mapped. It 

also does not exclude grazing, one of the principal 

causes of degradation in wetlands. 

The Threatened Species Conservation Act 1995 

requires environmental impact assessment for any 

developments which may have an impact on species 

of fauna or flora listed as either endangered or 

vulnerable in Schedules 1 and 2 of that Act. There is 

also provision for entire plant communities to be 

14 

listed as endangered, and for particular populations 

of species which may not be threatened at State level 

to be listed as locally threatened. Generally, the 

vegetation types which are most depleted and least 

reserved in the region do not contain any species of 

either flora or fauna which are listed as threatened, 

and hence the Act provides no protection currently 

for these vegetation types. Vegetation types of similar 

species composition and facing similar threats have 

been listed as endangered communities in the 

Sydney region, the Australian Capital Territory and 

Victoria. Nomination of some of the most depleted 

vegetation types in this region as endangered 

communities is an option which needs to be 

considered. Such listing has no impact on normal 

agricultural activities, but could be used to control 

potentially damaging developments, and might 

stimulate funding for further survey work. 

The Bega Valley Local Environmental Plan, 

administered by Council, provides another avenue 

for protection of significant vegetation. Techniques 

available to Council to improve protection of 

significant vegetation include: 

• extension of environmental protection zonings 

along riverine corridors; 

• definition of further areas of high scenic quality 

for scenic protection zoning; and, possibly 

• delineation on Local Environmental Plan maps 

of areas within the general agricultural zones 

where sensitive vegetation exists and where 

additional development assessment requirements 

might apply. 

A further technique for implementing the 

recommendations of this report could be the 

production of a Rural Lands Development Control 

Plan. This could specify requirements for 

subdivision and other development, and include 

incentives and other positive measures for 

landowners to guide development to those parts of 

their properties where impacts on native vegetation 

and habitat are less significant. 

Other incentive schemes which are currently in 

operation are the Native Vegetation Incentive Fund 

administered by the Department of Land and Water 

Conservation, and the Voluntary Conservation 

Agreements administered by the National Parks and


Wildlife Service. Both these schemes provide 

funding incentives (in the form of payment for 

fencing materials) for landholders to manage parts of 

their properties specifically for conservation. A 

contract between the parties is involved, which in 

the case of Voluntary Conservation Agreements 

results in a covenant being placed on the property 

title. Both these schemes, while admirable in intent, 

appear to be under-resourced in the region. There is 

also a shortage of expert advice available from 

extension staff for landholders wishing to pursue 

conservation of remnant vegetation under their 

control. 

Threatened flora in Bega Valley Shire 

Plant species listed as threatened in Commonwealth 

or State legislation occur primarily on public land 

tenures in the Shire. There is a cluster of species 

occurring on rhyolite outcrops, principally west of 

Pambula, of which several are listed as threatened 

and are endemic to Bega Valley Shire. Three of these 

(Zieria formosa, Z. parrisiae and Z. buxijugum) are 

known only from very restricted locations on private 

property, and require some additional form of 

protection such as from a Development Control 

Plan. The National Parks and Wildlife Service is 

currently preparing a Recovery Plan for these and 

some other rhyolite species in consultation with 

SFNSW and relevant landholders. 

The summit of Dr George Mountain is another 

threatened species hot spot, with several species 

occurring at a site which has received a high level of 

disturbance. It appears that the site may be a 

recreation reserve under the control of the Council 

(D Hamer, DLWC, pers. comm.), and it is desirable 

that a management plan be produced in 

consultation with the National Parks and Wildlife 

Service and other authorities which use the site. 

Species that are formally listed as threatened or rare, 

and which occur in Bega Valley Shire, are listed in 

Appendix 5 (p. 133). Species whose only occurrence 

in the Shire is within Wadbilliga National Park have 

been excluded. 

In addition to the species formally recognised as rare 

or threatened, there are a considerable number of 

plant species found in Bega Valley Shire which are 

listed as being uncommon within the Eden 

Management Area, and a few which are uncommon 

throughout their distribution (Keith & Ashby 1992; 

Keith, Miles & McKenzie 1999). Some of these are 

known from only one record in the region, and may 

in fact be more threatened in the area than plants 

which are formally recognised as threatened. A high 

proportion of the listed threatened species occur 

entirely or largely within National Parks. Many of 

the uncommon species do also, but quite a high 

proportion of them are found only or mainly on 

private property, where it can be assumed their 

abundance has been reduced by clearing or 

agricultural activities. Some of these species are more 

common in other parts of their range, but some, 

because of a habitat preference for areas which have 

been most affected by agriculture, are uncommon or 

declining throughout their range. 

Species which are of regional conservation 

significance because they are regionally uncommon, 

or because their habitat has been depleted by 

clearing and agriculture, are listed in Appendix 6 

(p. 136). 

Recommendations 

1. In general, vegetation types which occur outside 

the agricultural areas of the Shire are adequately 

represented in conservation reserves. The highest 

priority for conservation of vegetation in the 

region on private property should therefore be 

given to those types listed in Table 2 (p. 11). The 

Native Vegetation Incentive Fund and Voluntary 

Conservation Agreements are two vehicles 

currently available for doing this. Within each 

vegetation type of conservation significance, sites 

with remnant vegetation will vary in quality. 

Factors which will assist in determining the value 

of a site for conservation are outlined in points 1 

to 8 earlier in this summary. 

2. The exception to Recommendation 1 would be 

when the vegetation in question has a high 

strategic importance as a wildlife corridor. In this 

case the vegetation type would be of less concern, 

though the quality of the site would still be 

important, since a degraded site may be less 

capable of supporting fauna. The site would need 

to be considered in the context of surrounding 

vegetation and its probable persistence in the 

15


16 

long term. Factors such as use of nearby forest by 

threatened fauna species would also be 

important. 

3. The location of potential wildlife corridor areas 

is described in Part B, section 7.1. Council 

should require development applications for 

activities in these areas to include specific 

consideration of the habitat values for fauna. The 

Department of Land and Water Conservation 

should scrutinise clearing applications in these 

areas closely. In order to make this 

recommendation workable it will be necessary to 

more rigorously define the wildlife corridors. 

This may be a task for the Regional Native 

Vegetation Committee (when appointed). 

4. The Regional Native Vegetation Committee may 

also need to assess the value of current vegetation 

protection legislation for the most depleted 

vegetation types in the region. In this respect, it 

would be valuable to nominate some of the 

vegetation types listed in Table 2 (p. 11) as 

endangered communities under the Threatened 

Species Conservation Act 1995. 

5. Council could produce a Development Control 

Plan to assist with regulating development in 

areas which predominantly support vegetation 

belonging to the types listed in Table 2. 

However, it should be remembered that the CRA 

vegetation map is not accurate in fine detail, and 

so provides only a rough guide as to the areas 

which might be of significance. 

6. Council’s Tree Preservation Order should be 

reviewed to determine whether it can be used to 

protect remnant vegetation. Extension of the 

Tree Preservation Order to cover rural residential 

lots within agricultural zonings could be 

valuable. 

7. Council needs to assess the conservation 

significance of areas of land which are under its 

direct control, such as roadsides, cemeteries, 

showgrounds and other such crown reserves. 

Plans of management for such sites are needed. 

The reserve on Dr George Mountain is a special 

case requiring urgent action because of the high 

proportion of rare or threatened species on the 

site. 

8. The Rural Lands Protection Board needs to 

assess the travelling stock reserves under its 

control. Depending on past management 

practices, these reserves may be in better 

condition than surrounding private lands. Many 

are located within the core agricultural areas 

where significant vegetation types might be 

expected to occur. 

9. Council may be able to ameliorate the impacts of 

medium density residential development by 

attention to design features of subdivisions. It is 

desirable to minimise fragmentation of 

remaining vegetation, since this leaves it 

vulnerable to degradation by weed invasion and 

bell miner colonies. Control of run-off to 

prevent nutrient-enriched storm water entering 

remnant vegetation is important. The use of ring 

roads between housing and remaining bush will 

provide fire protection for houses and may 

reduce public pressure to burn adjacent bush too 

frequently. It may also help reduce the amount of 

garden waste dumped in nearby bush, which 

contributes to weed invasion. 

10. Council and the Far South Coast Catchment 

Management Committee should continue to 

encourage community groups to undertake 

vegetation restoration projects on both public 

and private land. However, these groups need to 

be given adequate access to professional advice to 

ensure that their activities are appropriate. The 

appointment of extension staff in the region with 

vegetation management as a primary focus needs 

to be given a high priority. Funding applications 

for on-ground works need to be scrutinised to 

ensure that planting proposals are suitable for the 

site. Community groups should be discouraged 

from planting on sites where natural 

regeneration may be just as effective. Control of 

environmental weeds to facilitate natural 

regeneration needs to be given a higher profile in 

the region. 

11. Provision of incentives for conservation on 

private land needs to be maintained and better 

resourced.


12. Control of willows needs particular attention 

because of the impact they can have on 

streambank stability and the scale of the seedling 

problem in the Shire. Addressing this problem is 

beyond the ability of community groups, 

although some groups are doing work on a 

limited scale. There needs to be development 

and implementation of a detailed willow control 

plan, and employment of a crew to control 

willows in accordance with the plan. Priority 

needs to be given to removal in areas where 

willows are still uncommon, and where the bulk 

of the vegetation is native, particularly in 

upstream sections of the river systems. 

13. When providing advice to landholders on 

plantation establishment, the Farm Forestry 

Project needs to take into account the possible 

impacts of planting exotic or non-local species 

(in terms of the spread of environmental weeds) 

and local species of non-local provenance. Rather 

than encouraging the use of exotic and non-local 

species, more consideration needs to be given to 

the use of suitable local species and the means by 

which their performance can be improved. The 

siting of plantations close to high 

conservation-value remnant vegetation requires 

careful consideration of potential impacts. 

14. Similarly, community groups, Council staff and 

the Roads and Traffic Authority should be 

discouraged from making inappropriate 

plantings of non-local species on public land. 

Council should consider the production of a 

booklet for distribution to rate-payers which 

discusses native vegetation conservation, and 

specifically lists environmental weeds which 

should not be planted. 

15. Because of the fine-scale inaccuracies in the CRA 

vegetation map, it will still be necessary, when 

considering vegetation management issues at the 

individual property scale, to do field inspections 

of vegetation to determine whether or not it falls 

into a significant vegetation type. To refine the 

map to the point where it is 100 per cent reliable 

would be a massive undertaking. It is probably 

therefore more efficient to provide training to 

staff in the region who may need to recognise 

relevant vegetation types (for example, Council, 

DLWC, NPWS). Council ground staff 

responsible for the management of roadsides and 

reserves, and voluntary community groups such 

as those managing cemeteries, may also need 

such training. 

16. A survey of the extent and impact of bell miner 

dieback across all land tenures in the region is 

urgently required. 

17

Assessment of data currently 

available on vegetation in 


Part A


Map 1: Study area with land tenure as at 1 January 1999 

20


1 Introduction 

This report documents the information currently 

available on vegetation in Bega Valley Shire. Part A 

assesses the accuracy and usefulness of some of this 

information. Part B is a report on the state of the 

vegetation. It summarises the conservation 

significance and status of the various vegetation 

types described for the district, discusses threats to 

vegetation and presents suggestions as to future 

action on vegetation issues. 

1.1 Comprehensive regional 

assessment data 

A wealth of information on vegetation within the 

Eden Native Forests Management Area (which 

includes Bega Valley Shire) has become available as a 

consequence of the recently negotiated Eden Region 

Forest Agreement for the Eden Native Forests 

Management Area. The Eden Region Forest 

Agreement (Anon 1999) has been negotiated 

between the State Government Ministers for Urban 

Affairs and Planning, Forestry, Environment and 

Fisheries, and outlines the terms under which forest 

management will be conducted in the region for the 

next 20 years. The Eden Region Forest Agreement 

was preceded and informed by a period of data 

gathering and analysis (the Comprehensive Regional 

Assessment, or CRA). The CRA included collation 

and collection of data on flora and fauna 

distributions. It resulted in the production of a 

model of vegetation communities and their 

distribution within the area prior to European 

settlement (the CRA pre-1750 vegetation map) and 

currently (the extant vegetation map). Comparison 

of these two modelled vegetation maps provides for 

estimates of the amount of clearing of particular 

vegetation types, and can be overlaid on land tenure 

maps to give an estimate of the degree to which each 

vegetation type is represented in local conservation 

reserves and on other tenures. The pre-1750 and 

extant vegetation maps are available as digital layers 

on ArcView, held locally by Merimbula NPWS, 

State Forests of NSW, the South East Forests 

Conservation Council and Bega Valley Shire 

Council. 

Other data provided as a result of the CRA include 

digital maps of forest growth stage showing 

distribution of ‘old growth’ forest and regrowth 

resulting from integrated logging or wildfire. The 

latter was derived from aerial photo interpretation, 

as was a map of broad forest classes based on the tree 

species present in the forest canopy. Unfortunately 

the broad forest classes do not generally coincide 

with the vegetation types used in the CRA-modelled 

vegetation map, so there are in fact two alternative 

vegetation maps available for the area, whose 

accuracy it is difficult to compare. 

Also available are a number of digital maps which are 

derived from various spatial data layers, sometimes 

including the CRA vegetation map, and represent 

attempts to model fauna habitat in the region. These 

are partly based on particular features of the 

vegetation such as the presence of a shrubby (as 

opposed to a grassy) forest understorey, or the likely 

presence of tree species known to provide particular 

resources needed by certain fauna species. The 

assessment of these maps was outside the scope of 

this project. 

The CRA vegetation map comes with a considerable 

amount of written documentation (Keith & 

Bedward 1998; Keith & Bedward 1999). In addition 

there are papers (Keith & Sanders 1990; Keith 

1996) and reports (Keith 1994; Keith, Bedward & 

Smith 1995) describing earlier phases of this work 

which may assist with the interpretation of the 

vegetation map. Most importantly, the reports by 

Keith and Bedward discuss the conservation 

significance of the different vegetation types, that is, 

the degree of clearing and reservation of each type in 

the region, and whether the type is unique to the 

region or more widely occurring in surrounding 

areas. 

Supporting information on the Broad Forest Class 

Map derived from air photo interpretation is less 

accessible. Only a draft report (Anon 1998) was 

available at the time this project was in progress. 

This report states that the air photo interpretation 

forest class mapping is not intended as an alternative 

to the CRA vegetation map derived from modelling, 

21


and that overall accuracy of the final map has not 

been subjected to either field validation 

(‘ground-truthing’) or desktop validation (unlike the 

vegetation modelling). In the light of this opinion, 

and because the Broad Forest Class Map does not 

come with supporting information about the 

conservation significance of the vegetation types it 

uses, it was decided to do little assessment of this 

product. Additional supporting documentation has 

subsequently become available in the form of the 

Field Guide to the South East Forests of New South 

Wales (SFNSW 1998), which includes a description 

of the eucalypt forests of the Eden Native Forests 

Management Area. The forests are described as 26 

forest ecosystem types within five ecosystem groups, 

and each type is correlated with one or more of the 

Keith and Bedward vegetation types and Research 

Note 17 (FCNSW 1989) forest types. A map 

indicates the approximate distribution of each type. 

Research Note 17 itself describes 232 vegetation types 

found in New South Wales, of which perhaps 40 

eucalypt forest types and 14 non-eucalypt forest 

types are relevant to the Eden Native Forests 

Management Area. 

The modelled CRA vegetation map represents a 

considerable advance on previously available 

information. Prior to this information being released 

into the public domain by the Resource and 

Conservation Assessment Council, there was no 

mapping of vegetation types available for the region. 

Earlier reports (Keith & Sanders 1990; Keith, 

Bedward & Smith 1995) had outlined a vegetation 

classification scheme for the district essentially the 

same, although less detailed, than that described by 

Keith & Bedward (1998, 1999), but mapping was 

not available. State Forests of New South Wales 

(SFNSW) had an alternative vegetation classification 

system based on canopy species only, derived from 

Research Note 17, and omitting a number of 

non-forest vegetation types such as heaths and 

wetlands (Appendix 2, Eden EIS, Harris-Daishowa 

1986). Mapping of some State Forests from air 

photo interpretation by this system had been done, 

but the information was not publicly available. 

As a result of the CRA the vegetation of the Eden 

region has been surveyed more thoroughly than any 

other part of New South Wales, with the exception 

of the area around Sydney (Keith & Bedward 1999). 

22 

The availability of this information in the public 

arena provides an opportunity for Bega Valley Shire 

Council and the Far South Coast Catchment 

Management Committee (CMC) to make decisions 

on questions of land use and resource allocation on a 

better informed basis than was possible previously. It 

will also be valuable in constructing a Regional 

Vegetation Management Plan as required under the 

Native Vegetation Conservation Act 1997. 

1.2 Non-CRA sources of 

information on vegetation 

The appointment of the Far South Coast CMC in 

1994 provided the stimulus for a great deal of data 

collection and collation in the region. Projects 

initiated or encouraged by the CMC have included: 

• collation of existing data on water quality (Clark 

undated) and the collection of data for water 

quality snapshots within the Towamba 

catchment (Turner, Lambert & Dawson 1996) 

and other major catchments (Turner, Lambert & 

Dawson 1997); 

• the South Coast Land Degradation Survey, 

involving mapping from air photo interpretation 

of erosion and potential erosion sites (still in 

progress); 

• an analysis of stream flows in the Bemboka River 

catchment and the possible impacts of logging 

on water yields (Gutteridge Haskins & Davey 

Pty Ltd 1997); 

• a considerable volume of work on the 

geomorphology of some of the far south coast 

catchments by staff and students of Macquarie 

University (Brooks 1994; Sinai 1995; Fryirs 

1995; Brierley & Murn 1995; Brierley, Fryirs & 

Cohen 1996; Fryirs & Brierley 1998a) 

culminating in recommendations for prioritising 

riparian rehabilitation works in the Bega–Brogo 

catchment (Fryirs & Brierley 1998b); and 

• a survey of willows and willow seedlings within 

the major streams of the Shire (Riddell 1997). 

Within this period there has also been a State-wide 

reassessment of water allocation practices by the 

Department of Land and Water Conservation


(DLWC), necessitating the collection of baseline 

ecological data to inform decisions about 

environmental flows (AWT EnSight 1997). A survey 

of soils of the Bega–Goalen Point 1:100 000 map 

sheet, which covers about half of the Shire, has been 

completed (Tulau 1997). There have been a number 

of consultant’s reports provided to Bega Valley Shire 

Council on estuaries and coastal wetlands (for 

example, Patterson, Britton & Partners 1996; 

Felton-Taylor 1998). 

Few of these reports provide much direct 

information about vegetation, though some do 

describe the vegetation of particular areas (for 

example, Patterson, Britton & Partners for Wallaga 

Lake and Felton-Taylor for Pambula racecourse 

wetlands). Riparian (streamside) vegetation has been 

surveyed quite intensively in the district, but this 

unfortunately did not result in the mapping of 

native vegetation. The willow mapping (Riddell 

1997), although it maps willow distribution in great 

detail, gives no information about native riparian 

vegetation. A map (Figure 19 in Fryirs & Brierley 

1998b) classifies reaches along streams of the 

Brogo–Bega catchment by degree of degradation and 

recovery potential, and this can be used to give some 

indication of which sections still carry riparian 

vegetation in reasonable condition (Fryirs, pers. 

comm.), but no detailed information has been 

recorded. A survey of wetlands in the Bega River 

catchment conducted by DLWC (Green 1999) 

identifies the location of remnant wetlands in 

agricultural areas but provides little information 

about the nature and condition of the wetland 

vegetation. 

Another field in which information is available is the 

identity and distribution of threatened plant species 

within the Shire. A digital database is maintained by 

NPWS giving approximate locations of records of 

species listed in Schedule 1 (endangered) and 

Schedule 2 (vulnerable) of the Threatened Species 

Conservation Act 1995 (the TSC Act). Council has 

access to this database under a licensing agreement 

and receives periodic updates. Keith and Ashby 

(1992) provide a list of species considered to be of 

conservation significance in the region. As well as 

species listed under the TSC Act, this report includes 

those listed as nationally rare (Briggs & Leigh 1988) 

but not threatened in New South Wales, those which 

are regionally uncommon or uncommon throughout 

their distribution and many which reach their limit 

of distribution in the district. It provides 

information about the locations at which the species 

were recorded. Additional species of significance 

recorded in the district during later surveys of 

remnant vegetation in the Bega valley are listed in 

Keith (1996). There is an updated listing of all 

vascular plants recorded for the Eden Native Forests 

Management Area (Keith, Miles & McKenzie 

1999), which identifies species of conservation 

significance, but does not provide information about 

their distribution. A list of significant species derived 

from this paper is presented in Appendix 5 (p. 133). 

23


2 Assessment of accuracy and 

usefulness of CRA-modelled 

vegetation distribution map 

One of the principal aims of this project is to 

determine to what extent the vegetation modelling 

conducted as part of the CRA for the Regional 

Forest Agreement process will provide a useful 

resource for Bega Valley Shire Council and the Far 

South Coast Catchment Management Committee. 

Potential uses include informing Local Environment 

Plans and a Regional Vegetation Management Plan, 

and indicating where resources can best be directed 

to address issues of land degradation and loss of 

biodiversity. 

There are two questions to be answered about the 

vegetation model described by Keith and Bedward 

(1998, 1999). First, how accurately do the 

vegetation types described in their reports reflect the 

species assemblages found in the district? Or, to put 

it another way, how easy is it to recognise the 

vegetation types in the field, from the information 

provided in the report? These two questions are not 

necessarily the same, since it is possible that the 

vegetation types described are accurate, but that the 

report does not provide sufficient information to 

permit easy recognition of them. Second, how 

successfully does the vegetation modelling predict 

the vegetation type that will be found in any locality? 

These questions are addressed in sections 2.4 

and 2.8 below. 

The vegetation map produced for the Eden CRA 

was not produced by mapping occurrences of the 

various vegetation types solely from aerial photos. 

Important environmental variables at the location 

where each vegetation type was sampled were 

determined. This information was then used to 

predict where else in the district each type should 

occur. The success of this method depends on the 

adequacy of sampling, on how well the crucial 

environmental variables have been selected from a 

large range of often very subtle variables which 

influence plant distribution, and on how accurately 

the environmental data layers used in the computer 

program reflect reality. Given that some of the 

24 

environmental data layers have also been modelled 

rather than mapped, there is scope for inaccuracy in 

this factor. Climatic variables could not have been 

obtained in any other way, given the scarcity of 

weather recording stations in the district. The 

authors state that while apparently accurate on a 

regional scale, the climate model may omit local 

effects such as frost hollows. The geology layer was 

derived from geological mapping originally 

conducted at a scale of 1:250 000. A map produced 

at this scale is unlikely to be totally accurate when 

used at the much more detailed scale of 1:100 000. 

In fact, errors are known to have occurred, and to 

have affected the accuracy of predicted vegetation 

locations. 

2.1 Derivation of the 

vegetation types 

The floristic data which contributed to the CRA 

vegetation map have been built up gradually over 10 

years, from 1987 to 1997. The data is based on a 

standardised method of data collection using 0.04 

hectare (generally 20 x 20 metre) plots. All plant 

species present in each plot were recorded and 

allocated a cover abundance score indicating their 

relative abundance on the plot, from ‘1’ (one to a 

few individuals and less than five per cent cover) to 

‘6’ (75–100 per cent cover). Information about 

vegetation structure (height and density of the tree, 

shrub and ground cover layers) and any disturbance 

was also recorded, along with environmental 

variables (rock and soil type present, altitude, aspect 

and the angle to the horizon at eight points of the 

compass, which is used to derive an exposure index 

for the site). 

Data collection was mostly done by NPWS staff and 

consultants, but where other vegetation surveys 

made use of the same or compatible methods, that 

data was incorporated as well.


To ensure that data were collected in a manner that 

adequately represented the range of environments 

present in the region, the area was stratified by 

geology, elevation and terrain. The stratification was 

refined over the 10-year period of data collection, 

with final sampling being targeted to fill gaps 

remaining from earlier field work. The final 

stratification was based on eight classes of geology, 

seven classes of altitude, two classes of slope and 

three classes of aspect. 

The first round of data collection provided 369 

sample plots spread broadly over the Eden 

Management Area, but excluding the privately 

owned cleared lands of the coastal valleys and the 

eastern Monaro tableland, and Wadbilliga National 

Park in the north-west of the area. The analysis of 

this data was reported in Keith and Sanders (1990) 

and produced a vegetation classification of the area 

with 41 vegetation types, of which three were further 

subdivided, giving 44 types in total. 

Over four years from 1992, further data were 

collected on private property, road verges and so on 

around the Bega valley and Quaama–Cobargo area 

to address the absence of information on the 

vegetation of agricultural areas. This was analysed 

separately and presented in Keith (1994) and Keith 

(1996), with 13 remnant vegetation types being 

described for these areas. A digital map was 

produced of remnant vegetation but did not receive 

very wide circulation. 

Re-analysis of the full data set in 1995 for the 

Interim Forest Agreement process used a data set of 

1066 plots (many more having been subsequently 

collected from public as well as privately owned 

land), and described 71 vegetation types. This 

information was made available in the form of a 

digital vegetation map and accompanying draft 

report (Keith, Bedward & Smith 1995). The need to 

fill gaps in coverage and the extension of the study 

area to include Wadbilliga and the Numeralla area 

generated another round of data collection, 

culminating in a final data set of 1680 plots. 

The latest data set was re-analysed in 1997. Of the 

614 new plots, only 406 could be assigned to an 

already described vegetation type. The 208 samples 

which could not be readily assigned to one of the 

already described vegetation types were discarded. It 

is not surprising that some samples could not be 

assigned to a vegetation type given that, in the field, 

areas of forest often appear as intermediate between 

two or more vegetation types, and that data were 

obtained from an area only 20 x 20 metres in size. 

The outcome of the new analysis was largely to 

maintain the existing 71 vegetation types. However, 

three types were split (46A and 46B, 22A and B, 

23A and B, the latter four occurring only on the 

Monaro tableland). Another six (W1 to 6) were 

added. These were known only from Wadbilliga 

National Park. Wadbilliga had not been included in 

the earlier analysis and data from this area were less 

complete, so that vegetation types W1 to W6 are 

regarded as interim. The same numbering system 

was maintained for the vegetation types to avoid 

confusion between the 1995 and 1997 versions of 

the vegetation map, but the names have changed in 

some instances. Earlier vegetation type names had 

included names of indicator plant species, for 

example ‘Lowland Dry Shrub Forest (Bloodwood)’. 

This was discontinued because it was found that 

people using the classification tended to look solely 

for the indicator species featured in the title, 

generally a tree, and failed to recognise the 

vegetation type if this species was not present. 

The analysis of the raw vegetation data into 

vegetation types was done by cluster analysis, using 

PATN software (Belbin 1994). This software 

clustered samples according to their degree of 

similarity. The first cut of the data produced 33 

groups. Some of the 33 groups were not split any 

further, while others could be further separated into 

more homogeneous species assemblages. These 

assemblages were found to correlate well with 

habitat differences (geology, topography, altitude) 

and, hence, were accepted as being real vegetation 

types. 

2.2 The vegetation types 

Keith and Bedward (1998) describe a total of 79 

vegetation types for the Eden Native Forest 

Management Area. This area includes some of the 

Southern Tableland, from Numeralla in the north to 

the Victorian border, with Bombala being on the 

western edge of the study area. A few of the 79 

vegetation types occur totally or largely outside Bega 

25


Valley Shire and are not covered in this report. The 

Eden Native Forest Management Area excludes a 

small area of Bega Valley Shire to the north and west 

of Cobargo, but vegetation types in this area are 

likely to be similar to those immediately to the 

south. 

Despite being referred to in the CRA report (Keith 

& Bedward 1998) as forest ecosystems, many of the 

vegetation types described are in fact not forests, 

being either coastal or estuarine communities, 

heaths, rock scrubs or wetlands. No native grasslands 

are described for Bega Valley Shire. There are a 

number of closed forest types, either rainforest or 

acacia scrub, and a total of 45 eucalypt-dominated 

forest types described for Bega Valley Shire. The 

latter can be broadly divided into: 

• wetter forests found mainly at higher altitudes on 

the escarpment; 

• intermediate forests found in gullies and on 

sheltered slopes at lower altitudes; 

• dry grassy forests found in the rain shadow 

valleys (Cobargo–Quaama, Bega and Towamba) 

on granitoid geology or in the granitoid 

hinterland south-west of Eden; and 

• dry shrub forests found generally on the less 

fertile soils and more exposed sites at all 

elevations throughout the Shire. 

Each vegetation type has been given a descriptive 

name which refers to the vegetation structure, nature 

of the understorey (grassy or shrubby), or a 

26 

particular or general location where it occurs. Each 

type also has a number, 1 to 70, and W1 to W6 for 

types found only in Wadbilliga National Park (which 

was added late to the study area). Vegetation types 

are referred to by number in this text, to save 

repeated use of rather lengthy names. Appendix 1 

(p. 100) provides a brief description of each 

vegetation type described for Bega Valley Shire. This 

appendix summarises the descriptions given in Keith 

and Bedward (1999) which provides a fuller written 

description of each vegetation type, diagnostic 

species and distribution. This information is 

presented only in the form of tables in Keith and 

Bedward (1998). The vegetation types are also listed 

in Table 4 (p. 35). 

2.3 Description of the 

vegetation types with 

indicator species tables 

Having arrived at a number of vegetation types, it is 

then necessary to describe them in such a way as to 

make them recognisable in the field. Because the 

CRA classification is based on full floristics rather 

than just the dominant tree species, as earlier 

classifications have been (Eden woodchip EIS, 

Harris-Daishowa 1986), it is necessary to be able to 

recognise a wide range of plant species in order to 

use the classification. 

Each vegetation type is described by means of several 

tables. An example is reproduced at Table 3.


Table 3: Sample diagnostic species description 

Species richness: 28 ± 5 (0.04 hectares) Extant area: 42 hectares 

Proportion cleared: 11 per cent (probably an underestimate) Number of samples: 6 

Diagnostic plant species of Map Unit 1, Dry Rainforest 

Species Target 

frequency 

Target C/A Residual 

frequency 

Residual C/A Fidelity class 

Brachychiton populneus 0.833 2 (1–3) 0.02 1 (1–1) positive 

Celastrus australis 0.833 3 (2–3) 0.008 1 (1–1) positive 

Ficus rubiginosa 1 5.5 (5–6) 0.005 1.5 (1–4) positive 

Geitonoplesium cymosum 1 2 (2–2) 0.106 1 (1–1) positive 

Notelaea venosa 0.5 3 (1–3) 0.112 1 (1–2) positive 

Pellaea falcata var falcata 1 2 (2–3) 0.094 1 (1–2) positive 

Pittosporum undulatum 0.833 3 (2–3) 0.089 1 (1–2) positive 

Plectranthus graveolens 0.5 2 (1–3) 0.009 1 (1–2) positive 

Sigesbeckia orientalis 0.5 2 (1–2) 0.076 1 (1–2) positive 

Solanum aviculare 0.167 2 (2–2) 0 0 (0–0) positive 

Alectryon subcinereus 0.667 1 (1–3) 0.005 1 (1–2) frequent 

Asplenium flabellifolium 0.833 1 (1–2) 0.119 1 (1–2) frequent 

Clematis glycinoides 0.667 1.5 (1–3) 0.113 1 (1–2) frequent 

Dichondra repens 0.833 1 (1–1) 0.282 2 (1–2) frequent 

Hymenanthera dentata 1 1 (1–2) 0.066 1 (1–2) frequent 

Oplismenus imbecillis 0.667 1.5 (1–2) 0.106 2 (1–2) frequent 

Urtica incisa 0.667 1.5 (1–2) 0.067 1 (1–2) frequent 

Eucalyptus polyanthemos 

ssp. vestita 

0.167 3 (3–3) 0.012 2 (2–3) uninformative 

Eucalyptus sieberi 0.167 1 (1–1) 0.283 3 (2–4) uninformative 

Eucalyptus tereticornis 0.333 2 (1–3) 0.046 3 (1–3) uninformative 

Lomandra longifolia 0 0 (0–0) 0.551 2 (1–2) negative 

Poa meionectes 0 0 (0–0) 0.541 2 (1–3) negative 

Pteridium esculentum 0 0 (0–0) 0.579 2 (1–3) negative 

Vegetation structure of Map Unit 1 

Frequency is the proportion of samples in which strata were present. Height and cover data are means with 

standard errors in parentheses (n = 6). 

Stratum Frequency (%) Height (m) Cover (%) 

Tree 50 23.0 (2.4) 6 (2) 

Small tree 100 10.8 (0.7) 80 (5) 

Shrub 100 1.7 (0.3) 16 (6) 

Ground cover 100 0.3 (0.0) 8 (2) 

27


Data presented for each vegetation type, in addition 

to diagnostic plant species, include: 

• the average number of species per sample plot 

(species richness); 

• the estimated area of this vegetation type 

currently present in the Eden Management Area 

(hectares); 

• the proportion estimated to have been cleared 

(per cent); 

• the number of samples which come up in the 

PATN analysis as belonging to this vegetation 

type; 

• a description of the general structure of the 

vegetation type in terms of what percentage of 

plots have the four layers (tree, small tree, shrub 

and ground cover) present, and the average 

height and percentage cover on the plots of each 

layer; 

• an indication of where in the landscape the 

vegetation type can be expected to occur 

(altitude, slope, aspect and geology on which 

plots occurred); and 

• a map of the Eden Management Area showing 

the location of the sample plots belonging to that 

type (not the modelled distribution of the 

vegetation type). 

The diagnostic species table lists a number of species 

which were found in samples of the particular 

vegetation type under discussion (the ‘target 

vegetation type’), though by no means all the species 

found in all samples from that vegetation type. Each 

listed species is assigned a fidelity class label: 

‘positive’, ‘frequent’, ‘uninformative’ or ‘negative’. 

Species labelled positive are those which occurred in 

half or more of the target vegetation type plots and 

were reasonably abundant in them (that is, cover 

abundance was greater than one: uncommon and 

less than five per cent cover), while being less 

abundant than this in the residual plots. Residual 

plots are all the non-target vegetation type plots 

combined. However, species which were less 

abundant than this in the target plots could still be 

defined as positive if they never occurred in the 

residual plots. 

28 

Species are labelled frequent if they were found in 

more than half of the target plots and were 

reasonably abundant in them, but also occurred at a 

similar frequency in the residual plots. These are the 

common and widespread species that are 

encountered nearly everywhere, and are 

consequently not very informative about vegetation 

type. Three of the most ubiquitous species, bracken 

(Pteridium esculentum), Lomandra longifolia and the 

grass Poa meionectes, are presented in every table 

regardless of whether they occurred in the target 

vegetation type or not. If they did not occur in any 

of the target samples, or occurred much less 

frequently than in the residual plots, these three are 

labelled negative indicators. Otherwise they are 

labelled frequent. 

Species are labelled uninformative if they occurred in 

less than half of the target plots, or had a cover 

abundance score of one (few individuals and less 

than five per cent cover), and the same in the 

residual plots. Interpreted literally, this could mean 

that a species which turns up in 49 per cent of target 

plots and only one per cent of residual plots is still 

labelled uninformative. In the tables this label has 

mostly been applied to tree species, because all trees 

which were recorded in plots of the target vegetation 

type have been listed in the diagnostic species table 

regardless of their frequency of occurrence in that 

vegetation type. This has been done because trees are 

generally the most obvious plants in the landscape, 

and the ones which people attempting to use the 

classification are most likely to be able to identify. 

Shrubs and smaller plants, whose target frequency 

score suggests they are uninformative, have generally 

been omitted from the table. 

Target and residual frequency have been expressed as 

a proportion of one. For example, a target frequency 

of one indicates the plant species occurred in all 

target plots, while a frequency of 0.25 or 0.5 

indicates it was in one-quarter, or half, of the target 

plots.


2.4 Ease of recognition in 

the field of the vegetation 

types 

Some of the 79 described vegetation types are so 

distinctive as to be immediately obvious in the field. 

Examples are Cool Temperate Rainforest (8), 

Montane Heath (53), Riverine Forest (40) and 

Beach Strand Grassland (62). These are generally 

vegetation types which are associated with some 

extreme habitat which has a fairly well defined 

boundary, which keeps the vegetation type distinct 

from those surrounding it. 

Some are reasonably obvious where they occur in a 

‘pure’ form, and some can be quite difficult to 

recognise. It should be remembered that vegetation 

varies along a continuum from one ‘pure’ type to 

another, in response to environmental variables 

(temperature and moisture regimes, soil fertility and 

so on) which also operate in a continuous rather 

than stepped fashion. It would be unrealistic to 

expect every sample of vegetation to fall neatly into 

one of the vegetation types, since many will be 

intermediate between two or more adjacent types. 

For example, in a hilly coastal landscape, one would 

expect to find a wetter vegetation type, say 13 

(Hinterland Wet Fern Forest), in the gullies and a 

drier vegetation type, perhaps 32 (Coastal Foothills 

Dry Shrub Forest) on the ridge tops. In between on 

the slopes will be a zone of overlap between the two 

types. Here the typical species of both types may 

intermingle, where at the extremes (gully and ridge 

top) they separate out into obviously different 

species assemblages. For this reason it may be 

necessary to walk around an area sampling the 

extremes of habitat present where the vegetation 

types segregate better. This can provide an indication 

of which vegetation types are combining to make up 

the less readily identifiable species assemblage which 

may occur on intermediate sites. 

There are, however, a number of vegetation types 

which are difficult to distinguish, not because they 

merge into adjacent vegetation types, but because 

there is a great deal of species overlap between them. 

Examples are 6 and 7, Coast and Hinterland Warm 

Temperate Rainforest. The main tree species in both 

of these tends to be lilly pilly (Acmena smithii), often 

with Doryphora sassafras in the northern end of the 

Shire. A number of less common species provide the 

indicator species for discriminating these two types 

from each other, and if these are not present (and 

being less common, they may well not be) then 

separation may be difficult. Generally the two types 

are shown as occurring in different areas. However, 

there are some locations where the model shows 

both occurring (for example, at Yurammie, 

Murrabrine and Wadbilliga), with 7 mapped at 

higher elevations, sometimes within the same gully 

system as 6. 

The worst case of similarity between vegetation types 

is that of the dry shrub forests found on dry sites 

such as ridge tops and exposed slopes on infertile 

soils. There are 10 vegetation types described 

(42–45, 47–50, W1 and W5), all of which have 

similar species present and similar structural features. 

They tend to be low in height, since they grow on 

infertile, dry, fire-prone sites, and to have a sparse 

sclerophyllous shrub layer and a very sparse ground 

cover. Silvertop ash (E. sieberi) and stringybarks 

(E. globoidea and E. agglomerata), and occasionally 

Angophora floribunda, are generally ubiquitous tree 

species in all these types. The shrub layer consists of 

a limited range of species, including Persoonia 

linearis, Acacia terminalis, Acacia obtusifolia, 

Oxylobium ilicifolium, Platysace lanceolata, 

Leucopogon lanceolatus, Lomatia ilicifolia, Epacris 

impressa and Monotoca scoparia. While most of these 

species tend to be present in many of these 

vegetation types, their proportions may vary 

between types. Types which are particularly difficult 

to separate are 48, 49 and W1. The degree of species 

overlap between these three types is very high. Some 

of the types are described as occurring only on 

particular geologies (Genoa sandstone for 50, 

Mumbulla Mountain tonalite for 48) or in particular 

locations (Wadbilliga for W1 and W5), which may 

help to identify them. However, 49 is also mapped as 

occurring extensively in Wadbilliga, so it would not 

be safe to rely entirely on such clues. It is often not 

possible to separate these particular vegetation types 

in the field using only the diagnostic species tables 

provided. 

It should be remembered that the diagnostic species 

tables were created to give some indication of what 

the characteristic species of each vegetation type are, 

29


rather than to provide a guide to enable 

identification in the field of all the vegetation types. 

Access to the full data matrix may be required to 

positively allocate a vegetation sample to a type. 

Even then it may not be possible to do so, given the 

variability of vegetation in the field discussed above. 

In other words, it may not be possible to do any 

better than to say vegetation is ‘49-ish’ at a site. 

Some species which appear to be quite good 

indicator species for some of the vegetation types, 

and which have been listed as such in earlier reports 

(Keith & Sanders 1990; Keith, Bedward & Smith 

1995) have been omitted from the diagnostic species 

lists in Keith and Bedward (1998). This is because 

analysis of a larger data set showed that they were 

less reliably attached to a particular vegetation type 

than they were initially thought to be. However, 

observation suggests that they were in fact useful 

indicator species and that some helpful information 

has been lost with their omission. 

At the other extreme from good diagnostic plants 

which have been omitted are species which have 

been included on the grounds of a single occurrence 

in the entire survey, that is, in one out of about 1600 

plots. These are presumably plants which are quite 

rare in the district, or belong to groups such as 

orchids, which flower only rarely and are invisible or 

unidentifiable when not flowering. Plants such as 

this are not useful as indicator species, partly because 

of the low probability of encountering them in the 

field, but also because one cannot be sure that they 

actually have any diagnostic power for the vegetation 

type they have been sampled in. Observations in 

some of the relevant vegetation types suggest that 

these plants are in fact not reliably associated with 

the types for which they have been listed as 

indicators. These species can be detected in the 

diagnostic species tables by their low target 

frequency. Dividing one by the number of samples 

of the target vegetation type will give the target 

frequency figure for which to look out. Residual 

frequency will be zero. In the sample diagnostic 

species description shown in Table 3 (p. 27) Solanum 

aviculare is such a species. 

When using the indicator species tables, care needs 

to be taken when applying the ‘uninformative’ label. 

Where a tree species is equally common in the target 

30 

and residual plots, its presence in the target plot is 

genuinely uninformative. However, because of the 

blanket application of the rule that anything under a 

frequency of 0.5 in both target and residual plots 

prompts a label of uninformative, some quite useful 

indicator species have been misleadingly labelled. 

For example, in vegetation type 19, Bega Wet Shrub 

Forest, blue box (Eucalyptus baueriana) is described 

as uninformative despite occurring with a target 

frequency of 0.368 and a residual frequency of only 

0.017. Another way of interpreting this is to say that 

had 1000 plots been done in type 19, 368 of them 

would have contained E. baueriana compared with 

only 17 of 1000 residual plots. This is surely a 

substantial enough difference to call E. baueriana a 

positive indicator species, even though it occurred in 

less than half of the target plots. So, when using the 

diagnostic species list, it is necessary to take note of 

target and residual frequency figures, and not just 

rely on the fidelity class label. 

To summarise the usefulness of the vegetation 

classification described by Keith and Bedward 

(1998, 1999), most of the vegetation types appear to 

be ‘real’ types which, with some practice, can be 

discriminated in the field, though only by an 

observer with a high level of botanical knowledge. 

The vegetation types fall into groups which occur in 

similar habitats and contain an often substantially 

overlapping species range. While the groupings are 

always easy to distinguish (for example, rain shadow 

valley dry grass forests dominated by forest red gum 

from dry shrub forests on poor soils dominated by 

silvertop ash or stringybarks), within each grouping 

the individual types may be more difficult. The 

vegetation types are not presented in groups in Keith 

and Bedward (1998), or at least the groups are not 

explicitly stated. The numbering system tends to 

place similar vegetation types together but there are 

some which do not fall in sequence. This report 

attempts to place them in a more logical order (see 

Appendix 1, p. 100). 

Types which are most difficult to discriminate 

between when using only the diagnostic species 

tables provided are: 

• the two warm temperate rainforests, 6 and 7; 

• several of the dry shrub silvertop or stringybark 

types (42–45, 47–50, W1, W5);


• 41 and W2 which are white ash-dominated 

forests of the top of the escarpment; 

• 9 and 10 (wet layered forests on the top of the 

escarpment); 

• 56 and 57 (wet heaths of the southern 

hinterland); 

• 54 and 55 (coastal heaths); and 

• 20 and 21 (red gum-dominated grassy forests of 

the Bega valley). 

It has been suggested that some of these types have 

been split too far and should be merged for ease of 

use, however the findings of this report do not 

endorse this view. It has been found that 

distinguishing between particular vegetation types 

becomes easier as familiarity with them increases. It 

may be a degree of inexperience with some of the 

vegetation types mentioned above which leads to the 

conclusion that they are difficult to tell apart. 

In the case of 20 and 21 (Bega and Candelo Dry 

Grass Forests), the difficulty is created by a poor 

choice of indicator species. Many of the grasses and 

herbs which are listed as diagnostic for 20 and 21 are 

present in both types. Some of the species which are 

most useful for discriminating 21 from 20 have been 

omitted from the tables. These are the shrub 

Jacksonia scoparia and the daisies Chrysocephalum 

apiculatum and Calotis lappulacea, usually found in 

21. These species tend to occur mostly on roadsides. 

Their infrequent occurrence in sample plots may be 

because of their palatability to livestock. 

In practice, the difficulty of distinguishing similar 

vegetation types in the field may not constitute a 

problem. This is because very similar vegetation 

types usually have similar conservation significance 

and, in most cases, this is the information that is of 

greatest concern. For example, silvertop and 

stringybark types are minimally cleared and 

adequately reserved, while red gum-dominated types 

(18, 20 and 21) are all highly cleared and 

unreserved. 

2.5 Vegetation types in 

which the range of species 

assemblages is inadequately 

described 

Some vegetation types could be split further. The 

fact that they have not is probably due to inadequate 

sampling, either of the vegetation type as a whole, or 

of the sub-type which occurs on a minority of sites. 

As an example of improved accuracy with increased 

sampling, a type called Estuarine Complex was 

described in Keith and Sanders (1990). Subsequent 

data collection has enabled the splitting of this type 

into four seagrass communities (67–70), two 

mangroves (65 and 66), a salt marsh (64) and a 

paperbark swamp (63) in Keith and Bedward 

(1998). 

Vegetation types which appear to include more than 

one species assemblage include: 

• 60 (Floodplain Wetland); 

• 19 (Bega Wet Shrub Forest); 

• 32 (Coastal Foothills Dry Shrub Forest) and 34 

(Coastal Gully Shrub Forest); 

• 38 and 39 (Riparian Scrubs); 

• 46B (Lowland Dry Shrub Forest); 

• 52 (Mountain Rock Scrub); 

• 62 (Coastal Scrub); and 

• 6 (Coastal Warm Temperate Rainforest). 

More detailed information is presented in 

Appendix 3 (p. 122). 

2.6 Derivation of the 

CRA-modelled vegetation 

maps 

The distribution of the 79 vegetation types was 

modelled using information gained during 

vegetation sampling to indicate in what sort of 

habitats the various vegetation types could be 

expected to occur. 

Digital data layers relating to features of the habitat 

were used to build up a picture of the terrain, 

31


geology and climate of the area. Topography was 

derived from a 25 metre grid digital elevation model 

obtained from the New South Wales Land 

Information Centre. This model included altitude, 

slope and aspect as well as more complex features 

such as: 

• a solar radiation index, giving an indication of 

site exposure; 

• a wetness index, representing the volume of 

water draining to a given point in the landscape; 

and 

• a neighbourhood topographic index, giving 

position in the local ridge and gully system. 

Temperature and rainfall variables (minimum 

temperature of the coldest month, maximum 

temperature of the warmest month, average annual 

rainfall, average monthly rainfall of the wettest and 

driest months) were modelled using ESOCLIM 

software (Hutchinson 1989). 

Geology layers were supplied by the Bureau of 

Resource Sciences. Their classification was modified 

so as to make some distinction between substrate 

types which were lacking: coastal sands from riverine 

alluvium, Lochiel basalt from other associated 

Devonian geologies and Genoa sandstone from 

other Devonian sediments. Final geological 

information was in the form of three data layers: a 

detailed one with 106 units, one with these lumped 

into 28 classes, and a broad classification into seven 

major formations (coastal sands, riverine alluvium, 

Tertiary alluvium, high quartz sediments, low quartz 

sediments, granitoids and basalt). 

A vegetation layer differentiating native vegetation 

from cleared land or plantations of exotics was 

prepared from a 1994 Landsat image (the ‘woody/ 

non-woody’ layer). This could be superimposed over 

the modelled pre-1750 vegetation distributions to 

produce a map of extant native vegetation. 

Another geographic information system (GIS) layer, 

differentiating eucalypt and various 

non-eucalypt-dominated vegetation types 

(rainforests, heaths, rock scrubs, riparian scrubs and 

estuarine communities), was prepared by manual 

interpretation of 1:25 000 scale black and white 

aerial photographs from 1963. These older photos 

were used because they are of better quality for 

interpretation than more recent photos and because 

32 

they pre-date integrated logging. Separate 

non-eucalypt forest mapping from 1994 colour 

aerial photos was prepared as part of the Eden CRA 

process, and this was used instead of the layer 

prepared earlier from black and white photos for 

mapping rainforests and acacia scrub, because it 

included Wadbilliga, which the earlier work had not 

done. 

Vegetation distribution was mapped by applying a 

set of decision rules about where each vegetation 

type would be expected to occur. The distribution of 

each vegetation type is described by between one and 

31 decision rules. The rules relate to the spatial 

features described above. Features most used in 

formulating the decision rules were: 

• broad geological class; 

• altitude; 

• average annual rainfall; 

• average rainfall of the driest and wettest months; 

• average maximum temperature of the warmest 

month; and 

• local topographic position. 

Slope and aspect were not often used. 

Some of the non-eucalypt vegetation types were 

mapped straight from aerial photo interpretation. 

These were two rainforests with very restricted 

distributions (1 and 5), Riverine Forest (40), Acacia 

Scrub (4) and the estuarine group (63, 64, 66). In 

the latter group, the seagrasses (67–70) have not 

been included in the CRA vegetation map, nor has 

the very restricted River Mangrove (65). 

Other non-eucalypt vegetation types were mapped 

primarily by aerial photo interpretation, with some 

interpretation by a small number of decision rules 

required to discriminate between closely related 

types. For example, the three more widespread 

rainforest types (6, 7 and 8) were distinguished on 

the basis of altitude, temperature and rainfall, and 

distance from the coast. The two riparian scrubs (38 

and 39) were mapped on the basis of rainfall, with 

38 (Southern Riparian Scrub) mapped as occurring 

in wetter areas. The rock scrubs (51 and 52) were 

discriminated on the basis of geology (rhyolite or 

granitoids) and the heaths (53–57) on altitude.


For areas which came up in the vegetation GIS layer 

as eucalypt-dominated, applying the decision rules 

to the spatial data produced a map of the presumed 

pre-1750 distribution of these vegetation types. The 

extant vegetation map was produced by overlaying 

the Landsat data on the location of clearing and 

plantations. 

2.7 Checks on map accuracy 

during production 

Qualitative checking of the accuracy of the 

modelling was conducted while the decision rules 

were being formulated. As vegetation types were 

mapped, their distribution was compared with the 

distribution of their sample plots, and compared 

with topographic maps to ensure that they were 

being mapped as occurring in the expected parts of 

the landscape (for example, in gullies or on ridges). 

If discrepancies were identified, the decision rules 

were modified to correct them. 

Quantitative checking was accomplished by 

withholding 154 (about 10 per cent) of the samples 

from the modelling process used to produce the 

map, and subsequently using these to check the map 

accuracy. That is, a sample from a particular location 

was checked against the map at that location to see 

whether the sample supported the prediction of 

which vegetation type would occur there. The 

samples used for checking came from vegetation 

types 13, 14, 15, 31, 32, 33, 34, 37, 42, 46A, 46B, 

47, 49, 53, 61, 62, W1 and W6. Accuracy was 

found to vary from 72 per cent within 100 metres to 

96 per cent within one kilometre. In other words, 

there was a 72 per cent chance that the sample 

vegetation type would be mapped as occurring 

within 100 metres of the exact spot from which the 

test sample came, but this improved to a 96 per cent 

chance that it would be mapped within one 

kilometre of the site. On a regional and local (within 

one kilometre) scale it appears that, at least for the 

vegetation types tested, the map is quite accurate. 

2.8 Assessment of map 

accuracy at fine scale 

Part of the brief of this project was to determine to 

what extent the CRA vegetation map is useful for 

making decisions relating to vegetation at a finer 

scale (that is, on particular sites). It is beyond the 

scope of the project to check the accuracy of the map 

for all vegetation types. Selection of which types to 

check was made on the basis of: 

• their rarity; 

• the degree to which they have been cleared or 

altered by human activity (farming, forestry, 

urban development); 

• the proportion of their original distribution 

thought to be represented in conservation 

reserves; and 

• the degree to which Council has any influence 

over their future use (that is, land tenure). 

Vegetation types which occur within the Eden 

Management Area but outside the Shire boundaries 

have been ignored. These are types 22A, 22B, 23A, 

23B, 24 and 25 (which fall entirely outside the Shire) 

and 50, 53 and 59 (which fall largely outside the 

Shire). 

Within the Shire, a number of vegetation types have 

their occurrence almost entirely within conservation 

reserves (national parks, nature reserves and flora 

reserves) and these will receive little further coverage 

in this report. These include: 

• some of the wet forest types from higher 

altitudes, which are now well reserved in the 

South East Forests National Park, Tantawangalo, 

Bemboka and Coolangubra Sections (types 8, 9, 

10, 11, 12, 15 and 16); 

• 41, which is a drier high altitude type reserved in 

similar locations; 

• 43, 44 and 45, which occur only in the 

south-west corner of the Shire and are largely 

contained within South East Forests National 

Park, Genoa or Coolangubra Sections; and 

• W1–W6, which occur largely within Wadbilliga 

and South East Forests National Parks (Bemboka 

Section). 

Also well reserved are: 

• 47 (in Ben Boyd National Park and Nadgee 

Nature Reserve); 

• 48 (in Biamanga National Park); 

• 28 (in South East Forests National Park, 

Coolangubra Section); 

33


• two rainforests with restricted distribution (2 

and 5 in South East Forests National Park, 

Coolangubra Section, and Mimosa Rocks 

National Parks respectively); 

• Rock Scrubs (3, 4, 51, 52 in South East Forests 

(Coolangubra), Wadbilliga, South East Forests 

(Yowaka) National Parks and Mount Poole Flora 

Reserve respectively); and 

• the heaths, 54 and 55 in Ben Boyd National Park 

and Nadgee Nature Reserve. 

Table 4 indicates which vegetation types have been 

most reduced in extent and are least well reserved. 

The most depleted and least well reserved vegetation 

types have been highlighted. It is these types on 

which the remainder of this report concentrates. 

Checking of map accuracy was confined to 

vegetation types which are thought to be 

significantly depleted from their pre-1750 extent, or 

are naturally rare in the Shire, and have been 

mapped as having a significant proportion of their 

occurrences on private land tenure. There are a 

number of naturally rare vegetation types for which 

a higher reservation target than the 15 per cent 

recommended in the JANIS criteria is considered 

desirable (JANIS 1996). However, since many of 

these are found primarily on public land tenures, no 

attempt has been made to assess the map accuracy 

for them. 

The CRA vegetation model appears to provide a 

reasonable guide to the vegetation types present and 

their distribution at a regional scale, inasmuch as the 

distribution of the vegetation types predicted 

corresponds reasonably well in broad terms to that 

which would be intuitively expected. At the local 

(one kilometre radius) scale, the map usually 

correctly predicts the main vegetation types present. 

This has been found to be the case in various coastal 

locations (most checking was done north of 

Pambula), and in many locations within the 

agricultural areas around Cobargo, the Bega valley 

and Towamba valley, which is where most checking 

was done. 

However, its accuracy at a fine scale is not such that 

it could be relied upon to indicate correctly the 

34 

vegetation present on any particular site. While it is 

generally (though not invariably) correct about the 

major vegetation types present within a one 

kilometre radius of any site, it tends to over-predict 

the occurrence of small patches of other vegetation 

types. This often appears to happen through 

overestimating the impact which landscape features 

such as minor gullies will have on the vegetation. 

Conversely, the model underestimates the effect of 

aspect on vegetation at fine scale, resulting in large 

areas appearing as uniformly of one vegetation type, 

when in fact there may be two or more vegetation 

types represented on different slopes of a hill or 

valley. The map is therefore frequently incorrect on 

the relative abundance of the different vegetation 

types predicted for a site. 

Particular sources of error are: 

• misinterpretation of vegetation types from aerial 

photo interpretation (for example, confounding 

of acacia scrub and rainforest, failure to detect 

casuarina forest against a background of eucalypt 

forest); 

• inaccuracies in the spatial data layers (climate, 

geology) which have been used to predict the 

location of the vegetation types (for example, 

poor separation of riverine alluvium and coastal 

sands in the geology layer leading to 

confounding of vegetation types predicted for 

sands such as 61 and 36, with those predicted for 

alluvium such as 60, 63 and 64); 

• over-mapping of some vegetation types resulting 

from failure to recognise the degree to which 

they are tied to particular landscape features (for 

example, 18 to north-facing slopes, 19 to 

drainage lines, particularly on steeper terrain); 

• inadequate sampling of some vegetation types, 

resulting in failure to adequately describe the 

type (for example, freshwater wetlands, of which 

only one sub-type, Melaleuca ericifolia scrub, has 

been sampled). 

A more detailed discussion of fine-scale accuracy for 

particular vegetation types is provided in Appendix 2 

(p. 111).


Table 4: Estimated area of vegetation types and proportion in reserves 

Vegetation type Pre-1750 

extent 

(ha) 

Extant 

area 

1997 

(ha) 

% 

cleared 

% 

reserved 

(NP, NR, 

FR) 

NP & NR 

(ha) 

Flora 

Reserves 

(ha) 

State 

Forest 

(ha) 

1 Dry Rainforest 47 42 11 64 30 0 0 9 

5 Bunga Head 

Rainforest 

6 Coastal Warm 

Temperate Rainforest 

7 Hinterland Warm 

Temperate Rainforest 

8 Cool Temp 


51 Rhyolite Rock 

Scrub 

52 Mountain Rock 

Scrub 

3 Rocky Top Dry 

Shrub Forest 

9 9 0 100 7 0 0 0 

Private 

land 

(ha) 

6 469 6 393 1 42 2 612 119 2 223 1 384 

3 053 3 027 1 55 1 603 64 591 702 

1 053 1 053 0 81 850 4 119 79 

51 51 0 80 22 19 10 0 

202 202 0 92 168 18 10 6 

1 188 1 188 0 84 995 0 166 22 

4 Acacia Scrub 6 673 6 288 6 54 3 617 0 16 2 644 

61 Coastal Scrub 2 273 1 505 34 50 1 128 0 4 222 

54 Mount Nadgee 

Heath 

55 Coastal Lowland 

Heath 

56 Hinterland 

Heath 

371 371 0 98 365 0 6 0 

1 676 1 630 3 89 1 490 0 37 93 

385 385 0 3 12 1 362 10 

57 Lowland Swamp 2 010 1 892 6 52 908 145 676 141 

63 Estuarine 

Wetland Scrub 

3 028 932 69 3 91 0 17 741 

64 Salt Marsh 370 296 20 13 47 0 3 129 

66 Estuarine 

Wetland (mangrove) 

67–70 Seagrass 

Meadows 

38 Southern 

Riparian Scrub 

39 Northern 


56 38 31 6 0 0 0 25 

403 403 0 ? 

611 516 16 22 128 5 197 178 

761 485 36 5 39 0 13 426 

40 Riverine Forest 81 65 19 0 0 0 0 65 

58 Swamp Forest 1 080 953 12 35 373 6 529 36 

60 Floodplain 

Wetlands 

9 421 3 281 65 3 296 0 240 2 417 

35


Table 4: Estimated area of vegetation types and proportion in reserves (continued) 


extent 

(ha) 

9 High Mtn Wet 

Layered Forest 

10 Mtn Wet 

Layered Forest 

11 Tantawangalo 

Wet Shrub Forest 

12 Mtn Wet Fern 

Forest 

15 Mountain Wet 

Herb Forest 

16 Basalt Wet Herb 

Forest 

17 Flats Wet Herb 

Forest 

W4 Kydra Flats 

Wet Forest 

13 Hinterland Wet 

Fern Forest 

14 Hinterland Wet 

Shrub Forest 

34 Coastal Gully 

Shrub Forest 

37 Lowland Gully 

Shrub Forest 

W6 Wadbilliga 

River Valley Forest 

18 Brogo Wet Vine 

Forest 

19 Bega Wet Shrub 

Forest 

20 Bega Dry Grass 

Forest 

21 Candelo Dry 

Grass Forest 

27 Waalimma Dry 

Grass Forest 

28 Wog Wog Dry 

Grass Forest 

29 Nalbaugh Dry 

Grass Forest 

36 

Extant 

area 

1997 

(ha) 

% 

cleared 

% 

reserved 

(NP, NR, 

FR) 

NP & NR 

(ha) 

Flora 

Reserves 

(ha) 

State 

Forest 

(ha) 

2 267 1 813 20 69 1 486 76 180 63 

20 033 17 940 10 45 9 059 28 5 436 2 982 

792 790 0 92 723 0 59 4 

2 302 2 259 2 65 1 476 17 683 62 

41 581 30 875 26 32 12 674 498 13 345 3 853 

14 904 12 209 18 19 2 764 35 3 207 4 964 

3 553 2 931 18 22 766 8 701 1 377 

3 501 3 214 8 72 2 536 0 111 343 

48 321 44 040 9 50 23 846 397 11 375 7 981 

27 004 25 882 4 39 7 707 90 13 925 3 633 

16 155 14 155 12 25 4 111 0 5 137 4 528 

16 153 15 147 6 31 4 770 285 8 173 1 722 

1 902 1 897 0 76 1 450 5 322 115 

7 850 4306 45 10 778 27 557 2929 

47 749 16 908 65 4 2 058 8 2 491 11 990 

31 952 3 809 88 0.5 159 0 72 3 512 

17 873 1 463 92 0.5 89 0 0 1 374 

1 324 1 324 0 22 294 0 1 031 0 

1 304 922 29 58 757 0 138 23 

Private 

land 

(ha) 

2 597 1 936 25 23 582 2 1 005 256




extent 

(ha) 

30 Wallagaraugh 

Dry Grass Forest 

31 Hinterland Dry 

Grass Forest 

35 Escarpment Dry 

Grass Forest 

36 Dune Dry Shrub 

Forest 

32 Coastal Foothills 

Dry Shrub Forest 

33 Coastal Range 

Dry Shrub Forest 

46A Timbillica Dry 

Shrub Forest 

46B Lowland Dry 

Shrub Forest 

47 Eden Dry Shrub 

Forest 

48 Mumbulla Dry 

Shrub Forest 

49 Coastal Dry 

Shrub Forest 

2 Myanba Euc/Fig 

Forest 

41 Mountain 

Intermediate Shrub 

Forest 

42 Inland 

Intermediate Shrub 

Forest 

43 Mountain 

Sandstone Shrub 

Forest 

44 Foothills Dry 

Shrub Forest 

45 Mountain Dry 

Shrub Forest 

50 Genoa Dry 

Shrub Forest 

Extant 

area 

1997 

(ha) 

% 

cleared 

% 

reserved 

(NP, NR, 

FR) 

NP & NR 

(ha) 

Flora 

Reserves 

(ha) 

State 

Forest 

(ha) 

1 663 914 45 16 273 0 400 228 

32 925 27 586 16 28 9 319 60 13 104 4 676 

34 577 22 007 36 19 6 231 251 3 731 10 840 

1 023 604 41 23 240 0 5 245 

24 521 23 401 5 24 5 919 41 11 956 4 441 

16 298 16 136 1 43 7 072 6 7 930 1 061 

22 917 22 792 1 8 1 164 610 20 497 497 

15 978 15 121 5 40 6 384 0 5 941 2 127 

17 797 17 141 4 66 11 727 108 4 098 965 

4 497 4 455 1 70 3 167 0 971 231 

32 334 31 837 2 33 6 739 794 21 042 3 150 

333 333 0 97 322 0 0 10 

1 865 1 864 0 74 1 361 22 418 54 

22 044 21 556 2 25 4 596 1 010 15 215 687 

2 492 2 479 1 89 2 229 0 96 154 

3 326 3 142 6 62 2 037 22 970 46 

Private 

land 

(ha) 

2 024 1 915 5 44 858 33 648 359 

3 702 3 026 18 55 1 996 42 776 200 

37




extent 

(ha) 

26 Tableland Dry 

Shrub Forest 

W1 Wadbilliga Dry 

Shrub Forest 

W2 Wadbilliga 

Range Shrub Forest 

W3 Wadbilliga 

Heath Forest 

W5 Wadbilliga 

Gorge Dry Forest 

Note: Highlighted vegetation types are those for which the proportion lost to clearing has been high, or the 

proportion in reserves falls below the relevant target level of 15, 60 or 100 per cent depending on the rarity of 

or degree of threat to remaining stands (JANIS 1996). Numbers are taken from Table 3.1 in Keith and 

Bedward (1 998) and are derived from the CRA vegetation model and extant vegetation map. However, this 

state of the vegetation report shows that the CRA vegetation model and extant vegetation map may 

overestimate or underestimate the area of the vegetation types which it assesses in detail, and presumably 

therefore of others as well. The figures should not be taken as being highly accurate. For example, the 

proportion of vegetation types 1, 18, 20, 21 and 36 in reserves is exaggerated. Appendix 2 (p. 111) provides 

detailed information for some of the highlighted vegetation types. 

38 

Extant 

area 

1997 

(ha) 

% 

cleared 

% 

reserved 

(NP, NR, 

FR) 

NP & NR 

(ha) 

Flora 

Reserves 

(ha) 

State 

Forest 

(ha) 

28 047 16 115 43 15 4 170 77 8 170 3 298 

27 352 27 341 0 98 26 747 0 237 205 

1 007 1 007 0 100 1 007 0 0 0 

3 085 3 085 0 99 3 060 0 0 24 

Private 

land 

(ha) 

7 748 7 239 7 71 5 461 3 930 823


3 Comparison of aerial photo 

interpretation vegetation typing 

with CRA vegetation model 

3.1 Derivation of aerial 

photo interpretation map 

As previously described, aerial photo interpretation 

(API) of vegetation was used to produce a GIS layer 

showing the location of non-eucalypt-dominated 

vegetation types, to feed into the Eden CRA 

vegetation map. Some non-eucalypt vegetation types 

were mapped directly from API, while others (for 

example, rainforests, heaths) required the use of 

decision rules to split them into two or more types 

which could not be distinguished from each other by 

API. 

The main task of the API mapping project, apart 

from mapping non-eucalypt vegetation types, was to 

provide information about forest growth stage, 

disturbance history from fire or logging and ‘old 

growth’ distribution. However, it was decided that 

API should assign eucalypt forests to broad forest 

classes, as well as mapping growth stage information, 

with the result that there are now two alternative 

digital vegetation maps available for the region. The 

pre-1750 vegetation map and extant vegetation map 

are based on models of vegetation distribution 

derived from ground-based sampling using the full 

range of plant species present, with some input from 

API with respect to the distribution of non-eucalypt 

forest vegetation types. The API map is derived 

directly and solely from interpretation of aerial 

photos and vegetation is classified only on the basis 

of the canopy species composition, that is (in the 

case of forests), the trees. 

The broad forest classes used for API mapping are 

derived from forest types described for New South 

Wales in Research Note 17 (FCNSW 1989). There 

are three levels of vegetation type identification: 

• Level 1 distinguishes eucalypt forest or woodland 

from rainforest or non-forest communities (for 

example, heath, wetland communities); 

• Level 2 delineates broad tree assemblages such as 

silvertop ash, stringybark or bloodwood 

complex, or forest red gum complex; 

• Level 3 is more detailed with, for example, forest 

red gum complex being further divided into 

Eucalyptus tereticornis–E. melliodora–E. maidenii, 

E. tereticornis–Angophora floribunda and 

E. tereticornis groups. Unfortunately, the Level 3 

groups only occasionally correspond with 

vegetation types used by Keith and Bedward 

(1998, 1999) for the CRA vegetation map. 

The CRA API draft report (Anon 1998) contains a 

table (Appendix 10.3) that indicates to which Keith 

and Bedward vegetation type and RN17 forest type 

each Level 3 group corresponds. However, numerous 

Level 3 groups do not directly correspond to types 

described in either of the other two classifications. In 

some instances the correspondence suggested in the 

API report is somewhat dubious. For example, 

relating E. tereticornis–Angophora floribunda to Keith 

and Bedward’s type 21, and E. tereticornis to type 20, 

when in fact both types 20 and 21 usually include 

both Eucalyptus tereticornis (forest red gum) and 

Angophora floribunda (apple). In several cases, the 

suggested correspondence is patently incorrect, as for 

example that between the Level 3 group Dry Heath 

Shrubland – Melaleuca armillaris – and the Keith 

and Bedward type 59, Subalpine Bog. In other cases, 

an obvious correspondence between a Level 3 group 

and a Keith and Bedward type has been omitted 

from the table, such as that between the group 

E baueriana (E. cypellocarpa–E. elata–E. viminalis– 

E. globoidea) and Keith and Bedward’s type 19, Bega 

Wet Shrub Forest. 

39


The Field Guide to the South East Forests of New 

South Wales (SFNSW 1998) does nothing to 

decrease the confusion over nomenclature. This 

document includes a description of 26 eucalypt 

forest types organised into five broad groups similar 

to those used by Keith and Bedward (wet layered 

forests, moist forests, intermediate shrubby forests, 

dry shrubby forests and dry grassy forests). Each 

forest type is related to one or more Keith and 

Bedward types and RN17 types, but the correlations 

and forest type names are not consistent between the 

Field Guide and the API Level 3 mapping groups. 

For example, in the Field Guide ‘Southern Blue 

Gum’ is equated to the Keith and Bedward type 13 

(Hinterland Wet Fern Forest) and is described as 

containing Eucalyptus cypellocarpa, E. muelleriana, 

E. obliqua and E. fastigata. The API report describes 

the ‘Southern Blue Gum Complex’ as containing 

three groups at Level 3, E. pseudoglobulus, 

E. maidenii and E. maidenii–E. globoidea. The latter 

of these three groups is supposedly equivalent to 

Keith and Bedward’s types 28 and 35, and for the 

others equivalents are not given. ‘Southern Blue 

Gum’ therefore refers to completely different forest 

types in the Field Guide and the API report (and 

hence API Broad Forest Class map). 

The conclusion is inescapable that the 

documentation available to assist with interpretation 

of the API vegetation map lacks internal consistency, 

making it very difficult to use. Debate is possible 

over whether a map based on API and intuitively 

derived vegetation types provides a more or less 

accurate representation of vegetation distribution in 

the district than one based on a model derived from 

ground-based sampling and computer analysis of 

species associations. However, the ultimate basis for 

the decision to use one or other product is likely to 

be made on the basis of their ‘user-friendliness’. 

Judged on this criterion, the Keith and Bedward 

vegetation maps and their supporting 

documentation are a clearly superior product. They 

provide a stand-alone, internally consistent product, 

which provides information about conservation 

significance of the vegetation types they describe. 

The latter is of particular importance in being able 

to apply the maps to making land use decisions, and 

is not provided with the API map. 

40 

3.2 Comparison of API and 

CRA vegetation map for 

specific locations 

Despite the above criticisms, it was decided to assess 

the relative merits of the two vegetation maps in a 

particular location. The area used was the 120 

hectare block owned by the Australian Bush 

Heritage Fund at Brogo. 

The API map and the CRA vegetation map were 

found to be equally inaccurate at fine scale, though 

in different ways. The CRA vegetation map shows 

most of the block carrying Bega Wet Shrub Forest 

(19), with smaller areas of Brogo Wet Vine Forest 

(18) on the tops of hills, and a few very small patches 

of Bega Dry Grass Forest (20) scattered apparently in 

gullies. Broadly, this is a reversal of the actual 

situation, as 18 is the more common vegetation type 

on the block. 

Both maps show Acacia Scrub and Riparian Scrub in 

the same locations, since the CRA vegetation map 

takes these vegetation types straight from API. Both 

maps fail to show three other vegetation types 

derived from API – Riverine Forest (40), Dry 

Rainforest (1) and Warm Temperate Rainforest (6 or 

7) – all of which occur in small stands on the block. 

However, given the small size of the stands, this 

omission is not surprising. 

The API map shows a type which is equivalent to 19 

(called River Peppermint-Apple complex – E. elata 

and Angophora floribunda) in one major gully system 

and on a steep south-facing slope, which is correct. 

However the API map fails to detect this type on the 

banks of the Brogo River and in three tributary 

gullies draining off a north-facing slope into the 

river. So, where the CRA map overstates the amount 

of 19, the API map understates it. The API map 

completely fails to detect any red gum forest types 

(18 or 20), despite 18 being the most common 

vegetation type on the block. However, the API map 

does detect the presence of a vegetation type which 

the CRA map does not. This is described as 

Gum-Box-Ironbark Complex and can include 

E. bosistoana, E. maidenii and E. globoidea. This 

assemblage certainly occurs fairly extensively on the


block and is similar to Escarpment Dry Grass Forest 

(35), which the CRA map does not show occurring 

in Brogo. 

Of the two maps, the CRA map is probably the 

more useful on this site in that it predicts the 

occurrence of three vegetation types of local 

conservation significance (18, 19 and 20) on the 

block, even if it does not correctly map their 

distribution. The API map only predicts one (19), 

and that in fairly restricted locations. The CRA map 

does a better job of signalling the conservation 

significance of the area. 

41


4 Conclusions 

The CRA vegetation map represents a considerable 

advance in knowledge of the vegetation of Bega 

Valley Shire. It describes vegetation in terms of full 

floristics rather than just the trees, thereby providing 

a much higher level of information about local 

vegetation than previous classification systems. It 

appears to describe vegetation types that are mostly 

real, and that can generally be recognised in the field 

(with some practice) by people with a reasonable 

degree of local botanical knowledge. However, 

because of the variability of vegetation in the field at 

a specific site, it may not be possible to allocate the 

vegetation to a type with certainty. This is partly 

because of the limited length of the diagnostic 

species tables provided (Keith and Bedward 1998, 

1999), but more due to the inherent variability of 

vegetation types in response to site factors. In 

practical terms this may not constitute a problem if 

all of the candidate vegetation types are of the same 

degree of conservation significance, which will 

generally be the case. 

In general, the modelling of the distribution of the 

vegetation types appears quite accurate at the 

regional scale, and reasonably accurate at the local 

scale in that the major predicted vegetation types are 

generally present roughly where predicted. However, 

ground truthing of a limited number of vegetation 

types, particularly those most likely to occur on 

private property, suggested that the level of accuracy 

is not so high that one could assume the map would 

be correct on any given spot location. Proportions of 

the vegetation types were often found to be rather 

different on the ground from those mapped. 

42 

What the map does do is provide an indication of 

broad areas in which vegetation types of greatest 

conservation significance are likely to occur, 

providing useful information for future vegetation 

management for Council, the Catchment 

Management Committee, the Department of Land 

and Water Conservation and the Regional Native 

Vegetation Management Committee (when 

appointed). However, for planning decisions at the 

individual property scale, field surveys will still be 

required to determine what vegetation type is 

present on a particular site. The Keith and Bedward 

classification does provide a framework in which to 

place observations of plant species. Site assessment 

may now be able to provide the name of the 

vegetation type occurring there (and some 

information about that vegetation type’s 

conservation significance) rather than a simple list of 

the plant species present on a site. 

The API Broad Forest Class map was not found to 

offer any advantages over the CRA vegetation map, 

and is considerably more difficult to interpret. 

Because the API map does take into account the 

effects of aspect on vegetation (unlike the CRA map) 

it may provide some useful additional information 

about vegetation distribution at particular locations. 

However, the general lack of correspondence 

between the broad forest classes used in the API 

mapping and the Keith and Bedward vegetation 

types does make use of the two maps confusing.

State of the vegetation report 

Part B


1 Introduction 

In addition to assessing the accuracy and usefulness 

of the vegetation mapping provided for the Eden 

Management Area, this report is intended to provide 

a summary of the information produced during the 

CRA process about the conservation status of the 

various vegetation types. 

Appendix 1 (p. 100) presents a brief description of 

the vegetation types described for Bega Valley Shire 

by Keith and Bedward (1998). Table 4 (p. 35) 

outlines the modelled pre-1750 extent and 

remaining extent of the vegetation types, and the 

percentage of their estimated pre-1750 extent 

thought to be included in the formal reserve system 

prior to the most recent round of National Park 

declarations in late 1998. 

While 79 vegetation types were described for the 

Eden Management Area, nine of these occur entirely 

or largely outside the boundaries of Bega Valley 

Shire. Although most of these are highly depleted by 

agricultural activities, and therefore are of high 

conservation significance, they are not discussed in 

this report. 

The remaining vegetation types can be grouped as 

follows. 

Dry Shrub Forests and Intermediate Shrub 

Forests – found from sea level to the top of the 

escarpment, usually on poorer soils derived from 

sedimentary rocks and often on ridges and exposed 

upper slopes. Most occur over a large aggregate area 

(up to 28 047 hectares pre-1750 extent) but some 

are more restricted. Those thought to have a 

naturally restricted distribution are: 

• 41, Mountain Intermediate Shrub Forest (1865 

hectares pre-1750 and current); and 

• W2, Wadbilliga Range Shrub Forest (1007 

hectares pre-1750 and current). 

Those with a more extensive distribution are: 

• 26, Tableland Dry Shrub Forest; 

• 32, Coastal Foothills Dry Shrub Forest; 

• 33, Coastal Range Dry Shrub Forest; 

44 

• 42, Inland Intermediate Shrub Forest; 

• 43, Mountain Sandstone Shrub Forest; 

• 44, Foothills Dry Shrub Forest; 

• 45, Mountain Dry Shrub Forest; 

• 46A, Timbillica Dry Shrub Forest; 

• 46B, Lowland Dry Shrub Forest; 

• 47, Eden Dry Shrub Forest; 

• 48, Mumbulla Dry Shrub Forest; 

• 49, Coastal Dry Shrub Forest; 

• 50, Genoa Dry Shrub Forest; 

• W1, Wadbilliga Dry Shrub Forest; 

• W3, Wadbilliga Heath Forest; and 

• W5, Wadbilliga Gorge Dry Forest. 

Wetter Eucalypt Forests – found either on the 

escarpment, or in sheltered situations in the coastal 

range and hinterland. Most of these are thought to 

have been quite extensive in their distributions with 

the following exceptions: 

• 9, High Mountain Wet Layered Forest (2267 

hectares pre-1750, 1813 hectares current); 

• 11, Tantawangalo Wet Shrub Forest (792 

hectares pre-1750 and current); 

• 12, Mountain Wet Fern Forest (2302 hectares 

pre-1750, 2259 hectares current); and 

• W6, Wadbilliga River Valley Forest (1902 

hectares pre-1750 and current). 

Types thought to be more extensive in their 

occurrence are: 

• 10, Mountain Wet Layered Forest; 

• 13, Hinterland Wet Fern Forest; 

• 14, Hinterland Wet Shrub Forest; 

• 15, Mountain Wet Herb Forest; 

• 16, Basalt Wet Herb Forest; and 

• W4, Kydra Flats Grass Forest.


Dry Grass Forests – found mostly on soils derived 

from granitoid rocks and in areas of relatively low 

relief in the dry rain shadow valleys north of Eden. 

Once extensive, these vegetation types have been 

largely cleared for agriculture. They are: 

• 18, Brogo Wet Vine Forest (7850 hectares 

pre-1750 to 4300 hectares current, 45 per cent 

cleared); 

• 19, Bega Wet Shrub Forest (47 749/16 908 

hectares, 65 per cent cleared); 

• 20, Bega Dry Grass Forest (31 952/3809 

hectares, 88 per cent cleared); and 

• 21, Candelo Dry Grass Forest (17 873/1463 

hectares, 92 per cent cleared). 

Type 19 could also be included with the wetter 

eucalypt forest types, but is placed with this group 

because of its similar distribution and clearing 

history. There are two other Dry Grass Forest types 

which are also extensive, and tend to occur on 

steeper sites, where they have been less affected by 

agricultural activities. These are: 

• 31, Hinterland Dry Grass Forest (32 925/27 586 

hectares, 16 per cent cleared); and 

• 35, Escarpment Dry Grass Forest (34 577/ 

22 007 hectares, 36 per cent cleared). 

Dry Grass Forests – found in the hinterland 

south-west of Eden, also on granitoids and often in 

areas of low relief. These have been less affected by 

clearing, though some areas thought to have carried 

these vegetation types are now under pine 

plantations. However, they are all naturally restricted 

in distribution. They are: 

• 27, Waalimma Dry Grass Forest (1324 hectares 

pre-1750 and current) 

• 28, Wog Wog Dry Grass Forest (1304 hectares 

pre-1750, 922 hectares current) 

• 29, Nalbaugh Dry Grass Forest (2597 hectares 

pre-1750, 1936 hectares current) 

• 30, Wallagaraugh Dry Grass Forest (1663 

hectares pre-1750, 914 hectares current). 

Vegetation types with naturally restricted 

distributions – related to particular topographic or 

geological features. Former distributions are thought 

to have ranged from as little as one hectare for 65 

(River Mangrove) up to 6469 hectares for 6 (Coastal 

Warm Temperate Rainforest), though the latter 

figure is probably an overestimate due to mapping 

errors. The features to which these vegetation types 

are confined include: 

• sheltered gullies (rainforests); 

• drainage lines (riparian scrubs and forests, wet 

heaths, wetlands); 

• rock outcrops (rock scrubs and some rainforests); 

• estuarine habitats (estuarine wetlands); 

• coastal sand deposits (beach strand and coastal 

dune scrub or forest); and 

• areas of shallow, very infertile or poorly drained 

soils (heaths). 

These vegetation types are listed below, with 

pre-1750 and estimated current area in parentheses 

for each type. 

• 1, Dry Rainforest (47/42 hectares); 

• 2, Myanba Eucalypt-Fig Forest (333/333 

hectares); 

• 5, Bunga Head Rainforest (9/9 hectares); 

• 6, Coastal Warm Temperate Rainforest (6469/ 

6393 hectares); 

• 7, Hinterland Warm Temperate Rainforest 

(3053/3027 hectares); 

• 8, Cool Temperate Rainforest (1053/1053 

hectares); 

• 38, Southern Riparian Scrub (611/516 hectares); 

• 39, Northern Riparian Scrub (761/485 

hectares); 

• 40, Riverine Forest (81/65 hectares); 

• 58, Swamp Forest (1080/953 hectares); 

• 17, Flats Wet Herb Forest (3553/2931 hectares); 

• 60, Floodplain Wetland (9421/3281 hectares); 

• 51, Rhyolite Rock Scrub (51/51 hectares); 

• 52, Mountain Rock Scrub (202/202 hectares); 

45


• 3, Rocky Tops Dry Shrub Forest (1188/1188 

hectares); 

• 4, Acacia Scrub (6673/6288 hectares); 

• 54, Mount Nadgee Heath 371/371 hectares); 

• 55, Coastal Lowland Heath (1676/1630 

hectares); 

• 56, Hinterland Heath (385/385 hectares); 

• 57, Lowland Swamp (2010/1892 hectares); 

• 63, Estuarine Wetland Scrub (3028/932 

hectares); 

• 64, Salt Marsh (370/296 hectares); 

• 65, River Mangrove (1/1 hectares); 

46 

• 66, Grey Mangrove (124/101 hectares); 

• 67–70, Seagrass Meadows (403/403 hectares); 

• 61, Coastal Scrub (2273/1505 hectares); 

• 62, Beach Strand Grassland (included with 

Coastal Scrub); and 

• 36, Dune Dry Shrub Forest (1023/604 hectares). 

It should be pointed out that some of the above 

estimates of past and current area of vegetation types 

are based on errors in mapping. For example, types 

18, 19, 36 and 4 appear to be rather less abundant 

than the figures indicate, while some others will 

presumably therefore be more abundant. The extent 

of rainforest in the district (types 6, 7 and 8) appears 

to have been exaggerated.


2 Conservation status of the 


2.1 Criteria for adequacy of 

conservation status 

Up until quite recently conservation reserves tended 

to be declared over areas which were perceived as 

being of no use for more productive purposes such as 

agriculture or forestry, or over areas with a high 

scenic value, the two features often coinciding. 

Historically, public forests have occupied areas with 

steeper topography and poorer soils which were 

passed over in the initial phase of clearing for 

agriculture. Those considered useful for timber 

production were progressively dedicated as State 

Forests while, after the National Parks and Wildlife 

Act 1967 was passed, areas with high scenic or 

recreational value were progressively added to the 

National Park estate. Increasing public concern for 

biodiversity issues has resulted in the transfer of 

substantial areas from State Forest to National Park 

in the Eden region in the past decade. It is only in 

the 1990s that there has been general acceptance of 

the need to reserve examples of all existing 

vegetation types (Austin & Margules 1986; 

Bedward, Pressey & Keith 1992a). Underlying the 

recent Eden Region Forest Agreement (EFA) is the 

premise that reserving samples of all vegetation types 

provides the most cost-effective way of conserving 

biodiversity of flora and fauna. Implementation of 

the National Forest Policy Statement 

(Commonwealth of Australia 1992) requires 

negotiation of Regional Forest Agreements between 

Commonwealth and State governments with the 

intended outcome being a comprehensive, adequate 

and representative reserve system and ecologically 

sustainable forest management for those areas 

remaining outside the reserve system. 

The question of what constitutes a comprehensive, 

adequate and representative reserve system has been 

addressed by the JANIS report (JANIS 1996). This 

report arrived at the figure of 15 per cent of the 

estimated extent of each vegetation type prior to 

European settlement as the minimum desirable 

reservation target. However the report also states 

that the greater the proportion of the vegetation type 

reserved, the higher the probability that it will 

continue to support viable populations of the flora 

and fauna dependent upon it in the long term. That 

is, greater than 15 per cent reservation is desirable 

where there is no land use conflict and more than 

15 per cent is achievable. Conversely, it may be 

appropriate in some instances to consider a figure of 

less than 15 per cent reservation where the relevant 

forest ecosystem is extensive in its distribution or 

where exploitative uses are of low intensity and the 

ecosystem has demonstrated resilience to this use. 

The reserve system should maximise the area of high 

quality habitat for all flora and fauna known to use 

it, but with particular reference to: 

• the needs of rare and threatened species; 

• organisms with special habitat requirements such 

as migratory species; and 

• areas of high species diversity, natural refugia for 

flora and fauna and areas with a high level of 

endemism (JANIS 1996). 

Reserves should be large enough to sustain viable 

populations in the long term. There needs to be 

replication across the full range of distribution of 

forest ecosystems to provide for conservation of 

genetic diversity and to provide some protection 

against loss of populations due to localised major 

events such as wildfire. The nature of the 

management of adjacent lands and their ability to 

provide corridors for genetic exchange and fauna 

movement is also important in developing a reserve 

system. 

Where forest ecosystems are fragmented and hence 

not amenable to full conservation within reserves, it 

may be necessary, where they occur on public lands, 

to manage them for conservation values by codes of 

practice or management plans. A system of informal 

reserves has been proposed under the EFA to provide 

conservation outcomes for some such vegetation 

47


types within State Forests. Where forest ecosystems 

occur solely or largely on privately owned lands, a 

range of incentive schemes may be used to achieve 

conservation outcomes. 

Where forest ecosystems are vulnerable or rare, a 

higher reservation rate than 15 per cent of original 

extent may be required. For ecosystems reduced to 

around 30 per cent of their original extent, or where 

there has been significant change in species 

composition or alteration to ecosystem processes, a 

reservation target of at least 60 per cent of remaining 

extent is desirable. Some ecosystems may be 

vulnerable even if not depleted. Examples would 

include naturally fragmented ecosystems which are 

sensitive to disturbance and have limited capacity to 

regenerate, such as some rainforests. Rainforests have 

in fact been targeted for 100 per cent reservation 

throughout the State regardless of their original 

extent, although this cannot be achieved by 

inclusion in formal reserves in the Eden region 

because of their naturally fragmented distribution. 

Where a forest ecosystem is rare or endangered it is 

desirable to reserve 100 per cent of existing 

occurrences where feasible. Rare ecosystems are 

those where geographic distribution is limited to a 

total range of less than 10 000 hectares, a total area 

of less than 1000 hectares or patch sizes of less than 

100 hectares, where these do not aggregate to 

significant areas. An endangered ecosystem is 

defined as one where the total area remaining is less 

than 10 per cent of its former area, or where 90 per 

cent of its area is in small patches which are subject 

to threatening processes which make them unlikely 

to persist in the long term (JANIS 1996). 

2.2 Conservation status of 

vegetation types in Bega 

Valley Shire 

With respect to the vegetation types (an equivalent 

term to JANIS’s ‘forest ecosystem’) found in Bega 

Valley Shire, many can be said to be adequately 

conserved in that at least 15 per cent of their 

presumed original extent is contained within the 

reserve system (assuming that the distribution 

predicted by the CRA vegetation model is correct). 

Generally, the types which are well conserved fall 

into two main groups – those which are found in 

extremely unproductive habitats and consequently 

48 

were placed in the reserve system quite early, and 

those which were transferred from State Forests to 

National Parks as a result of the Interim Forest 

Agreement (IFA). The IFA was the first stage of the 

implementation of the National Forest Policy 

Statement. The first group contains mostly Dry 

Shrub Forests which occur in Wadbilliga National 

Park and other steep, infertile areas within national 

parks, as well as coastal heaths reserved in Nadgee 

Nature Reserve and Ben Boyd National Park. The 

second group contains vegetation types which have 

their main occurrence along the escarpment, which 

are now reserved within the South East Forests 

National Park. There are also some very restricted 

vegetation types whose distribution happens to fall 

entirely within national parks. 

Restricted vegetation types which apparently have 

close to 100 per cent of their original extent reserved 

are: 

• Myanba Eucalypt-Fig Forest (2) and Rocky Tops 

Dry Shrub Forest (3) in South East Forests 

National Park (Coolangubra Section); 

• Bunga Head Rainforest (5) in Mimosa Rocks 

National Park; 

• Tantawangalo Shrub Forest (11) in South East 

Forests National Park (Tantawangalo Section); 

• Acacia Scrub (4) and the various types largely 

restricted to Wadbilliga (W2, W3, W4, W5, 

W6) in Wadbilliga and South East Forests 

National Park (Bemboka Section); 

• Mountain Rock Scrub (52) in Mount Poole 

Flora Reserve; and 

• some heaths (54 and 55) in Nadgee Nature 

Reserve and Ben Boyd National Park. 

More extensively occurring vegetation types which 

are reserved to the extent of 15 per cent or more of 

their estimated pre-1750 extent are: 

• the wetter escarpment eucalypt forest types 9, 

10, 12, 15 and 16; 

• the dry shrub forests 32, 33, 41–45, 46B, 47–50 

and W1; 

• wetter coastal vegetation types 13, 14, 34 and 

37; and 

• the dry grass forests of the hinterland 29, 31 and 

35.


The three most widespread rainforest types (6, 7 and 

8) are also reserved to more than 15 per cent (42 to 

81 per cent), as are the three restricted dry grass 

forests 27, 28 and 30 (16 to 58 per cent). 

Vegetation types classified as rare (see section 2.1 

above) are regarded as requiring 100 per cent 

reservation of remaining stands. However, these are 

often just the vegetation types that are difficult to 

place in the reserve system because of their naturally 

fragmented distribution. Examples are the rainforest 

types (6, 7 and 8) and other types also restricted 

largely to the vicinity of drainage lines such as 

Hinterland Heath (56), Swamp Forest (58) and 

Southern Riparian Scrub (38). As a result of the 

EFA, Southern Riparian Scrub has been protected 

by the creation of the Wallagaraugh Reserve along 

the Wallagaraugh River and Imlay Creek. It is 

intended that informal reserves within State Forests 

will provide for the conservation of types 56 and 58. 

Many of the drainage lines in the southern 

hinterland within Nadgee, Timbillica and Yambulla 

State Forests have been included in this network, 

which will be managed for conservation under 

approved management plans, in which logging is 

excluded (Land Tenure map accompanying Eden 

Region Forest Agreement, 1999). Rainforests will 

continue to be managed by prescription where they 

occur in production forests. That is, logging is 

excluded and a 10 metre buffer strip must be left 

around rainforest stands. 

Rhyolite Rock Scrub (51) is a rare and naturally 

fragmented vegetation type which is found on small 

rocky outcrops scattered within dry shrub forests. 

This type has 80 per cent of its occurrences within 

South East Forests National Park (Yowaka Section), 

Nethercote Falls and Jingera Flora Reserves. 

However, stands remain within Nullica State Forest, 

and on private property on the northern border of 

Jingera Flora Reserve. While 80 per cent would be a 

reasonable reservation level for many vegetation 

types with restricted distributions, 51 contains a very 

high proportion of rare, threatened or regionally 

endemic plants, which makes it more important to 

attempt to achieve the best possible management for 

all occurrences. A Recovery Plan is in preparation for 

the rhyolite endemics with the involvement of the 

NPWS, the SFNSW and some private landholders. 

The vegetation types which have been most depleted 

in the area are, for the most part, those which 

coincide in their distribution with agricultural areas. 

These are: 

• the four formerly widespread dry rain shadow 

valley vegetation types (18–21); 

• Northern Riparian Scrub (39); 

• Riverine Forest (40); 

• Floodplain Wetland (60); and 

• Dry Rainforest (1). 

Remaining stands of all these vegetation types occur 

principally on private land. The extent of their 

current reservation ranges from none to an estimated 

10 per cent for Brogo Wet Vine Forest (18), with the 

exception of Dry Rainforest (an estimated 64 per 

cent in South East Forests National Park 

(Coolangubra Section), though this figure is not 

accurate). 

The map on the following two pages (Map 2: 

Depleted vegetation types on private land) depicts 

the Eden Management Area and Bega Valley Shire, 

showing the location of woody vegetation (generally 

forest) on private land. 

The four most extensive vegetation types which are 

largely or entirely confined to private land (18–21) 

are mapped. Type 17, which has about half of its 

predicted occurrence on private land is also 

included. Less extensive vegetation types with a 

largely private land distribution such as Northern 

Riparian Scrub (39) and Riverine Forest (40) have 

been omitted, as occurrences are too small to be 

visible at this scale. 

Another common vegetation type of agricultural 

areas, acacia scrub, has been included, but it should 

be noted that the Acacia Scrub (4) mapped in 

Wadbilliga National Park and described in Keith and 

Bedward (1998, 1999) is a different plant 

association from that mapped on private land. The 

latter is almost all black wattle (Acacia mearnsii) 

regeneration, while vegetation type 4 describes an 

association typical of steep rocky slopes dominated 

by Acacia silvestris, such as occurs in Brogo Pass. 

49


Map 2: Depleted vegetation types on private land 

50


51


The Eden Region Forest Agreement (Anon 1999, 

Table 1, page 15) lists priorities for conservation in 

the region which cannot be achieved on public lands 

and these are reproduced in Table 5 below. Of the 

listed vegetation types in Table 5, the last two types 

in each column (including both low priority types) 

have all or nearly all of their private land occurrences 

outside Bega Valley Shire and are hence outside the 

scope of this report. Due to the errors which were 

found in mapping of the other vegetation types that 

do occur within Bega Valley Shire during field work 

for this report, some alterations in priority are 

suggested. With the exception of Flats Wet Herb 

Forest (17), which was not checked, nearly all of the 

listed vegetation types were found to have been 

overpredicted compared with their actual 

distribution. 

Most of the Bega Valley Shire vegetation types listed 

in Table 5 occur mainly or entirely in agricultural 

areas within the three rain shadow valleys (1, 18, 19, 

20, 21, 39, 40) while 36 is found only in coastal 

areas and 60 occurs in both coastal and agricultural 

areas. 

Dry Rainforest (1) has been described (Keith & 

Bedward 1998) as having 64 per cent of its 

occurrence in the Shire reserved (in South East 

Forests National Park, Coolangubra Section). 

However, this is an overestimate based on: 

• an exaggeration of the size of at least some of the 

stands within Coolangubra (through confusion 

of either Rocky Tops Dry Shrub Forest (3) or 

wattle regeneration with Dry Rainforest during 

air photo interpretation); and 

Table 5: Priority for protection of vegetation types 

52 

• failure to detect stands on private property 

around the margins of agricultural areas in the 

Bega valley. 

Stands on private property are very small, naturally 

fragmented and vulnerable to fire, stock browsing 

and trampling, and weed invasion. Reserved stands 

are also highly vulnerable to fire. This vegetation 

type should be accorded a high priority for 

protection on private property. It is represented on 

the 120 hectare property owned by the Australian 

Bush Heritage Fund at Brogo, which is managed for 

conservation, but the total area of Dry Rainforest on 

this block is quite small. Dry Rainforest stands 

which are reserved in Coolangubra appear to be 

quite species-poor by comparison with stands in the 

Brogo area on private property, with respect to vines 

and epiphytes. 

Brogo Wet Vine Forest (18) appears to have been 

substantially over-mapped on private property 

through failure to recognise the degree to which it is 

tied to particular topographic sites (upper 

north-facing slopes). It has also been mapped as 

having 10 per cent of its occurrence in reserves, 

principally South East Forests National Park 

(Bemboka Section) and Biamanga National Park 

along their boundaries with cleared land. This is 

definitely incorrect for Bemboka Section, and 

probably also for Biamanga. It is very unlikely that 

any Brogo Wet Vine Forest is reserved. It should 

therefore also be accorded a high priority. This type 

is extensive on the Australian Bush Heritage Fund 

property at Brogo. 

High priority Moderate priority Low priority 

19, Bega Wet Shrub Forest 1, Dry Rainforest 22A, Monaro Dry Grass Forest 

20, Bega Dry Grass Forest 17, Flats Wet Herb Forest 30, Wallagaraugh Dry Grass Forest 

21, Candelo Dry Grass Forest 18, Brogo Wet Vine Forest 

40, Riverine Forest 36, Dune Dry Shrub Forest 

60, Floodplain Wetland 39, Northern Riparian Scrub 

23A, Monaro Grassland 23B, Monaro Basalt Grass 

Woodland 

24, Subalpine Dry Shrub Forest 22B, Numeralla Dry Shrub 

Woodland


Threats to remaining stands on private property 

(inappropriate fire regimes, weeds, livestock and 

rabbit disturbance, and grazing) may be slightly 

lower than for other remnant vegetation types of 

agricultural areas because of the steep rocky terrain 

on which it typically occurs. 

Bega Wet Shrub Forest (19), although formerly 

widespread on lower-lying agricultural areas, also 

occurs on private land outside the main agricultural 

areas, such as on some coastal drainage lines and 

along the Burragate Road between Wyndham and 

Burragate. It was also, unlike types 18, 20 and 21, 

formerly common in the Towamba valley, though 

also extensively cleared there. This vegetation type’s 

reservation status will have improved due to the 

addition of Murrabrine State Forest to the reserve 

system and the addition of parts of Nullica State 

Forest to the South East Forests National Park as a 

result of the Eden Forest Agreement. However, as it 

had originally been over-mapped in these areas, 

particularly in steeper parts, it is likely that the bulk 

of the Bega Wet Shrub Forest in these areas will still 

be found on private property bordering the reserves, 

where the terrain is less steep. Small areas occur on 

the Australian Bush Heritage Fund block at Brogo. 

Based on the fact that some of this vegetation type is 

actually reserved, and that stands can be quite 

extensive, it is probably sufficient to give it a 

moderate priority for conservation on private lands. 

Bega Dry Grass Forest (20) and Candelo Dry Grass 

Forest (21) have been estimated to have only 0.5 per 

cent of their occurrence in reserves. In fact, none is 

reserved. Stands of these vegetation types are 

generally small, fragmented and subject to grazing, 

weed invasion and inappropriate fire regimes. They 

are often affected by loss of understorey, dieback and 

lack of regeneration. They are the most threatened 

vegetation types in the Shire. It is possible that the 

current total area of remnant stands has been 

estimated quite accurately, but many individual 

stands have probably been incorrectly identified as to 

type (for example, 21 mapped as 20, or 20 mapped 

as 18). Some small stands, such as those on road 

verges or stands with a low tree density, have not 

been mapped at all. These vegetation types should 

remain a high priority for protection on private 

lands and on suitable public lands such as road 

verges and travelling stock reserves. 

Northern Riparian Scrub (39) occurs principally 

within agricultural areas although there are minor, 

unmapped occurrences at the upper ends of rivers 

which drain out of reserves onto private property, 

such as Tantawangalo Creek. Mapping of this 

vegetation type on private property on the upper 

Towamba River is incorrect, as this area actually 

carries Southern Riparian Scrub (38) which is now 

well reserved in the Wallagaraugh catchment. 

Northern Riparian Scrub is in fact most common on 

private property along rocky sections of the 

Bemboka River and Tantawangalo Creek. The main 

threat to this vegetation type is weed invasion, 

particularly by blackberry, willows and African 

lovegrass, and the herbicide spraying that may be 

used to control these species. Northern Riparian 

Scrub should be given a high priority for protection, 

particularly as it protects stream banks from erosion 

as well as being of conservation significance as a 

severely depleted vegetation type. 

Riverine Forest (40) occurs only on private property 

in the Shire, with the exception of small, probably 

atypical stands in State Forest on the Murrah River. 

The dominant species, Casuarina cunninghamiana, 

reaches its southern limit of distribution around 

Bega, so the type is confined to streams north of this 

point (Narira Creek, Murrah River, Brogo River and 

some of their tributaries, lower Bega River). Some 

clearing has occurred (estimated at 20 per cent by 

Keith and Bedward, but probably greater) and many 

remaining stands are severely degraded by livestock 

trampling and browsing, lack of regeneration, weed 

invasion, erosion, altered hydrological regimes and 

occasionally severe mistletoe infestation. Original 

mapping of this vegetation type omitted several areas 

where it occurs, but Version 4 of the map includes 

all known stands in the Eden Management Area. 

Unmapped stands are likely to occur on the Yowrie 

River and Wandella Creek which are within the 

Shire boundary but outside the Eden Management 

Area. A high priority for protection of this 

vegetation type is supported. Like Northern 

Riparian Scrub it is also important in maintaining 

riverbank stability. 

Floodplain Wetland (60) is mapped as occurring 

primarily in coastal areas on the lower reaches of 

major rivers such as the Murrah, Bega and 

Towamba, as well as around smaller coastal drainage 

53


systems. Version 4 of the map extends mapping of 

this type into agricultural areas of the Bega valley, 

where small stands of either Melaleuca ericifolia 

scrub or reed beds and sedgelands remain in some 

uneroded drainage lines. Large areas of this 

vegetation type have been lost to erosion, or severely 

degraded by livestock trampling and grazing, and 

weed infestation. Deliberate clearing of Melaleuca 

scrub and draining of reed bed areas is also likely to 

have occurred. The type is naturally relatively rare in 

the district because of the short length of river 

systems and consequent limited area of floodplain 

available for the development of wetlands. Ground 

truthing of this vegetation type in coastal areas 

showed that where relatively large patches had been 

mapped, the type was actually confined to drainage 

lines within the mapped area, so total extent has 

been exaggerated. Very little (three per cent) is 

estimated to be reserved. A high priority for 

protection of this vegetation type is supported. 

Mapping is unlikely to have captured all 

occurrences, as they are generally small and difficult 

to detect from air photo interpretation. On small 

coastal drainage lines Melaleuca ericifolia would 

probably be concealed by surrounding eucalypt 

forest except where it has been retained in cleared 

areas. In agricultural areas, uneroded swampy 

drainage lines may stand out as being greener than 

surrounding pasture (or yellow, indicating 

Phragmites reed bed), but whether they consist of 

native wetland species, exotic pasture species or 

weeds could only be determined by site inspection. 

A survey of wetlands recently conducted by DLWC 

(Green 1999) used air photo interpretation to detect 

wetlands within the Bega–Brogo catchment. Field 

work was limited and the report provides little 

information about vegetation type or condition. The 

output of the report included a digital map of 

wetlands which may be an improvement on 

mapping of Floodplain Wetland for the CRA. 

Dune Dry Shrub Forest (36) is the only one of the 

priority vegetation types listed which is not found in 

agricultural areas. It is estimated to have been 

reduced to about 60 per cent of its former extent by 

clearing, mostly for coastal residential development, 

and to have 23 per cent of its former area in reserves. 

However, ground truthing for this vegetation type 

indicated that mapping is not very accurate, and that 

54 

most stands mapped within reserves either do not 

exist or are very much smaller than mapped. The 

main occurrence of this vegetation type in the Shire 

is along the coast immediately south of Bermagui 

and east of Main Road 272. Some of this area is 

vacant crown land and some is private property. The 

type formerly existed south of Merimbula Lake 

mouth, but has been almost entirely cleared for 

residential development in this area. Another 

relatively large area is located between Boydtown 

and the Nullica River mouth, south of Eden, 

although this stand consists of regrowth and is 

somewhat atypical in being dominated by Eucalyptus 

viminalis rather than E. botryoides. A small stand of 

similar species composition occurs on private 

property on the western shore of Wallagoot Lake. 

Threats to remaining stands include further clearing 

for development, inappropriate fire regimes and 

weed invasion (bitou bush, African lovegrass and 

bridal veil creeper being the most potentially 

damaging species). This vegetation type should be 

given a high priority for protection. 

The accuracy of mapping of Flats Wet Herb Forest 

(17) was not assessed for this report as the type did 

not appear to satisfy the criteria for being considered 

either rare or endangered. However, it has been listed 

as of moderate priority for protection in the EFA, 

presumably because of its naturally small, 

fragmented occurrences near drainage lines and the 

fact that half of its modelled occurrences are on 

private property. Most of these occurrences are 

predicted to be in the Wyndham area. 

Other vegetation types which should be given a 

moderate priority for conservation on private 

property are the warm temperate rainforest types (6 

and 7) and estuarine types (63–66). Rainforests, 

because of their naturally fragmented distribution, 

cannot be fully conserved in reserves. They are under 

threat in the region from inappropriate fire regimes, 

and on private property they may be degraded by 

livestock trampling and weed invasion. Damage by 

stock often seems to happen because the stock need 

to enter gullies carrying rainforest stands to find 

water. There are some high quality rainforest stands 

on agricultural land around the margins of the rain 

shadow valleys, and rainforest is present on private 

property along the coastal strip, where it may be 

subject to the same threats.


Rainforests are not shown as having been cleared to 

any appreciable extent by the vegetation model. 

However this may be as a result of mapping errors 

from air photo interpretation. It appeared in 

agricultural areas that wattle scrub regenerating 

downstream from remnant rainforest in gullies was 

often mapped as rainforest. This wattle regeneration 

may in fact have been occupying an area that had 

formerly carried rainforest, which had been cleared. 

The low clearing figures for rainforest do not take 

into account the loss of quality of rainforest stands as 

a result of wildfires such as the 1952 fire. Almost all 

rainforest stands in the district are regrowth, with 

old fire damage evident. Some stands which have 

been burnt more than once have degenerated to 

mere tangles of vines and may not survive another 

fire (Floyd 1982). 

Estuarine Wetland Scrub (63), Salt Marsh (64) and 

Mangrove (66) are estimated to have been quite 

extensively cleared for both agriculture (for example, 

parts of the shorelines of Wallaga, Wapengo and 

Merimbula Lakes) and for residential development. 

The proportion of these vegetation types in reserves 

is recorded as low, although it may have been 

underestimated. Mangroves particularly seem to 

have been under-mapped. Mangroves and seagrasses 

are protected from clearing under the Fisheries 

Management Act 1994, although the other estuarine 

vegetation types are not. Livestock trampling and 

browsing is probably the greatest threat to stands on 

private property. Weed invasion tends not to be high 

in the saline soils on which these vegetation types 

grow, although it may occur around the landward 

edges of Estuarine Wetland Scrub. 

55


3 Potential threats to vegetation 

and their impacts 

There are a number of threats to the long-term 

persistence and integrity of the under-reserved 

vegetation types in the Shire. The nature of the 

threats varies to some extent with land tenure, 

although many of the threats (such as altered fire 

regimes or weed invasion) operate across all tenures, 

though to differing degrees depending on the 

amount of human disturbance. In assessing the 

nature and level of threats to remnant vegetation, it 

seems more useful to contrast agricultural landscapes 

that have already been largely cleared with lands that 

are still largely forested, irrespective of land tenure. 

3.1 Agricultural landscapes 

In the dry rain shadow valleys of the 

Cobargo–Quaama area, the Bega valley, the 

Wyndham area, Towamba valley and, to a lesser 

extent, in parts of the coastal strip, much of the 

native vegetation has been cleared. Clearing appears 

to have been initially very thorough, extending into 

steep and unproductive marginal lands which have 

only regained some tree cover in the last few decades. 

Individual relict trees scattered in paddocks and 

along roadsides represent the only survivors from 

pre-European times, with virtually all the ‘remnant’ 

bush present today being regrowth. This bush varies 

greatly in quality, as measured by age, diversity of 

native species and degree of invasion by exotic 

species, depending on past management history. The 

impacts of clearing for agriculture have fallen 

disproportionately on the vegetation types found 

largely in the dry rain shadow valleys (types 18–21) 

as well as riparian and wetland types found in those 

areas (39, 40, 60), although other vegetation types 

would have been cleared for agriculture in coastal 

areas and around the valley margins. 

The land tenure within agricultural landscapes is 

almost entirely freehold private property. There are 

very few areas of leasehold land in Bega Valley Shire. 

Small areas of crown reserve land exist but these are 

often managed in a similar fashion to surrounding 

private land, making them of doubtful value for 

56 

vegetation conservation. Travelling stock reserves, 

showgrounds and cemeteries are examples. All of 

these are subject to either grazing (often quite 

intensive) or mowing (which may prevent 

regeneration of trees and shrubs and affect the 

species composition of the ground cover). In some 

parts of the State travelling stock reserves and 

cemeteries are recognised as providing valuable 

habitat for plants and animals which have been 

largely eliminated from adjacent grazing or cropping 

land. Such reserves in Bega Valley Shire require 

assessment to determine whether this is also the case 

here. Road verges are least likely to be subjected to 

grazing and they may support populations of plant 

species which are highly palatable to livestock and 

hence susceptible to elimination from adjacent 

private land. 

The main threats to remnant vegetation in 

agricultural areas are discussed below. 

3.1.1 Fragmentation and clearing 

Fragmentation of the remaining native vegetation 

into isolated, often small stands, which may not be 

capable of sustaining the necessary ecological 

processes for their long-term survival, has already 

largely happened in Bega Valley Shire. The current 

trend is now in the reverse direction, with 

regeneration having already occurred on a relatively 

broad scale in more marginal areas such as Brogo 

and Verona. On a small scale, regeneration continues 

on sites such as road verges, steep lands and some 

riparian areas where livestock access is limited. 

Nonetheless, clearing potentially remains a threat to 

remnant vegetation. Clearing of relatively large areas 

of original forest occurred as recently as the 1970s in 

Brogo, and clearing of smaller areas of regrowth 

forest continues to the present. This is now subject 

to some controls under the Native Vegetation 

Conservation Act 1997, and the impact of this will be 

considered in section 4.1. Ringbarking of regrowth 

and relict trees is still practised by some landowners 

in the district.


One vegetation type upon which clearing pressures 

have not eased in recent years is wetland vegetation 

found in drainage lines within agricultural areas. 

Many stands were destroyed during the early 

settlement phase by gully erosion or deliberate 

drainage and conversion to pasture, and cattle 

trampling and weed invasion are ongoing threats to 

remaining stands. A new threat has arisen recently – 

the construction of large dams for off-river water 

storage for irrigation. The difficulty of providing 

sufficient water resources for all irrigators from 

unregulated streams, and increased awareness of the 

need to maintain environmental flows, has led to 

greater emphasis on storing water in large dams. The 

construction of these could destroy some remaining 

areas of wetland vegetation. The damming of already 

eroded gullies is unlikely to have this effect, since 

wetland vegetation has already been lost from them. 

3.1.2 Dieback 

Dieback is the process of gradual loss of vigour, 

eventually leading to death of trees. It is a common 

feature of agricultural landscapes over much of the 

country. It presents a considerable threat to remnant 

vegetation, particularly that which consists of 

scattered old trees in pasture. 

The causes are numerous and interlinked. In 

remnant vegetation they may include: 

• the old age of relict trees; 

• drought stress; 

• soil salinisation and waterlogging; 

• defoliation by insects; 

• soil compaction and increased nutrient levels due 

to stock camping under trees; 

• damage to bark from stock rubbing or chewing; 

and 

• damage to bark and roots by agricultural 

machinery. 

The two most obvious immediate causes in 

agricultural areas on the far south coast are drought 

and the increased level of defoliation by Christmas 

beetles or scarabs (Anoplognathus spp.), with the two 

factors usually operating in tandem. Christmas 

beetles, having a larval stage which feeds on grass 

roots, are advantaged by increased pasture 

productivity, and can build up to higher population 

levels in areas with pasture-based agriculture. Recent 

reductions in the amount of ploughing in the district 

have probably decreased scarab mortality. Dry 

weather conditions appear also to decrease mortality. 

While drought is said to be a major cause of scarab 

larva mortality on the New South Wales tablelands, 

this factor may not operate in the same way on the 

coast, due to droughts being generally less severe, 

with less loss of vegetation cover and consequently 

less soil heating. Wet years appear to cause higher 

mortality, presumably due to fungus attack. 

Dry seasons also put the trees under more stress, 

resulting in the production of foliage with a higher 

nitrogen concentration (Marsh & Adams 1995; 

Landsberg & Wylie 1983). This more nutritious 

foliage enables scarabs and other leaf-eating insects 

to build up larger populations. In response to 

defoliation, eucalypts put on a flush of new juvenile 

foliage which is also higher in nitrogen than mature 

foliage, thus potentially creating a positive feedback 

loop, with repeated flushes of growth and 

defoliation. Eventually the tree’s reserves of 

carbohydrate are depleted and it may die. Generally 

on the far south coast, wetter conditions return and 

reduce insect pressure on trees, allowing some 

measure of recovery, but each drought event brings 

the district’s remnant trees closer to death and 

finishes off a few more. 

The noisy miner (Manorina melanocephala), an 

aggressive native honeyeater which defends 

communal territories against all other birds, may 

also be implicated in dieback in agricultural areas 

(Low 1994). Like its better-known relative the bell 

miner (M. melanophrys), the noisy miner feeds on 

insects, some of which are responsible for defoliation 

of eucalypts. By driving other birds out of its 

territory the noisy miner actually decreases predation 

on insects so that insect populations can increase to 

damaging levels. Experimental removal of noisy 

miner populations has been shown to result in an 

influx of other birds into the treated area although 

an improvement in tree health has not yet been 

demonstrated (Clark et al. 1995). Noisy miners 

prefer open woodland with a grassy understorey and 

little shrub cover, so remnant vegetation in grazed 

farmland is usually ideal habitat. Being honeyeaters, 

57


they also consume nectar when available, and 

plantings of flowering shrubs designed to attract 

birds may actually have the opposite effect, by 

improving the breeding success of noisy miners 

which then drive out other birds (Low 1994). 

Increasing rural residential development in the more 

open agricultural areas may therefore have the effect 

of exacerbating noisy miner-mediated dieback. 

In general, the most severe dieback is apparent in the 

drier centre of the Bega and Cobargo–Quaama 

valleys, where clearing has been most thorough and 

regeneration limited. Around the valley margins, and 

in the smaller Towamba valley, the effect is 

ameliorated by the presence of eucalypt 

regeneration. This makes the dieback less visually 

obvious, but probably also makes it less severe, by 

supporting higher levels of natural predators of 

defoliating insects. Also, where many trees are 

present the insects can spread their attention over 

many trees, rather than concentrating on, and 

therefore completely stripping, just a few. The valley 

margins are also likely to contain less productive 

pasture, being currently more utilised for rural 

residential than agricultural purposes, and hence 

support lower populations of scarab larvae. However, 

old relict trees are definitely also affected in these 

areas. 

Another potential cause of eucalypt deaths is high 

levels of infestation with mistletoe. While some 

degree of mistletoe infestation is a natural 

occurrence in forests, trees around forest edges and 

isolated trees can become heavily infested, with a 

detrimental effect on their health. Mistletoe levels 

seem to be higher in the Towamba valley than other 

agricultural areas of the Shire. This may be because 

of the relatively small area which has been cleared in 

the Towamba valley and the consequent close 

proximity of forest to all parts of the valley. The 

forest represents an original seed source of mistletoe, 

as well as providing habitat for the various bird 

species which spread the seeds. Mistletoe is 

controlled to some extent by fire, since the mistletoe 

plants are more likely to be killed by a 

moderate-to-hot fire than the tree which is their 

host. Since remnant vegetation within agricultural 

landscapes often has fire entirely excluded for long 

periods or is burnt only by controlled cool burns, 

this may provide a partial explanation for the 

increase in mistletoe populations. 

58 

It can be expected that the relict trees in agricultural 

areas which date from before European settlement 

will continue to die at an increasing rate in the 

future, since many must be reaching the end of their 

natural life span. This would not present a problem 

had adequate regeneration been occurring around 

them to provide a replacement generation. However 

in many areas this has not been the case, and the 

death of isolated paddock trees will leave some farms 

completely devoid of shade trees, with adverse 

consequences for livestock productivity, and for the 

general aesthetics in the district. In the more severely 

affected areas dieback is not confined to the oldest 

trees, with trees of only 70–80 years old suffering as 

badly, and regeneration struggling to survive. 

Apart from the problem of loss of shade trees for 

livestock, and the aesthetic effect, the loss of native 

trees and understorey on farms will also exacerbate 

the lack of natural insect predators which currently 

contributes to the problem. This could not only 

cause a worsening of the dieback, but make it more 

difficult to successfully establish tree plantations, and 

affect pasture productivity. The death of individual 

relict trees also represents the loss of an opportunity 

for natural regeneration. 

The loss of further tree cover will also contribute to 

rising water tables, with the possibility of dryland 

salinity occurring in some parts of the district. The 

Wolumla Creek catchment has been shown to be 

subject to a high degree of salinity of stream water 

compared with the Candelo and Tantawangalo 

catchments (Stahl 1997). Saline scalds have been 

observed in this area (N Stahl, pers. comm.). 

3.1.3 Loss of biodiversity – flora 

In terms of the total area of native vegetation in the 

Bega Valley Shire, there has been very substantial loss 

of biodiversity through clearing for agriculture. Trees 

and whatever shrubby understorey existed were 

removed over large areas and regeneration prevented 

by grazing pressure from livestock and rabbits. The 

native ground cover of grasses and herbs has been 

completely replaced in many areas by exotic grass 

and herb species (the latter mostly agricultural 

weeds). Cropping, principally of corn, peas and 

beans, which involved total removal of native 

vegetation, was also practiced in agricultural areas of 

the Shire until quite recently. Where native pastures 

persist they have been affected by grazing to varying


degrees, and also by changes to fire regimes. The 

probable impact on native pastures would involve 

loss of species which are highly palatable to stock or 

rabbits or unable to withstand constant grazing. An 

example is the replacement of kangaroo grass 

(Themeda australis) by weeping grass (Microlaena 

stipoides) in heavily grazed native pastures. Palatable 

herbs may also be lost, particularly those with erect 

growth habits, which are more susceptible to grazing 

than the more prostrate growth forms. 

Because there are no sites which are known to have 

been free of such disturbance with which to compare 

existing remnant vegetation, we have no way of 

judging the magnitude of species losses in the 

district. There has been a steady trickle of new 

records of plant species for the south coast from 

agricultural areas of the Shire in recent years, 

suggesting that some species are confined to this 

habitat within the district, and are very rare here, 

although they may be common in other parts of 

their range (for example, the herbs Bulbine bulbosa, 

Leptorhynchos squamatus and Lespedeza juncea). 

These may be species that were formerly more 

common in the Shire but have been almost 

eliminated, or they may be species for which Bega 

Valley Shire represents a marginal habitat and they 

were uncommon here before European settlement. 

The species diversity found within regenerating 

native vegetation varies depending on: 

• the degree of initial disturbance; 

• the length of time since clearing (which affects 

the number of species still present in the soil seed 

bank); 

• the presence of a seed source from remnant 

individual plants on the site; 

• the distance from other areas of native vegetation 

which might act as a seed source; 

• the past management of the land; and 

• the level of grazing pressure from livestock, 

rabbits and native herbivores, such as wallabies. 

Remnant vegetation may have a very diverse 

complement of native trees, shrubs and ground 

cover species, as is often the case around the valley 

margins. At the other extreme, it may consist of 

single species stands of young trees regenerating 

around a single relict tree above a ground cover 

consisting of a variable mix of native and exotic 

species. Early clearing of trees appears to have often 

been selective, with only one or two species being 

left as shade trees. Some farms can be seen to carry 

only forest red gum, while a neighbouring farm may 

carry only angophora. This practice will affect the 

tree species diversity which can occur in regeneration 

from site to site. Stringybarks are often eliminated 

from paddocks because of their susceptibility to 

ringbarking by stock rubbing or chewing on the 

bark, and so are probably under-represented in 

regeneration relative to their former abundance. 

It should be remembered that remnant vegetation 

need not contain any woody vegetation at all. The 

original understorey of Candelo Dry Grass Forest 

(21) and Bega Dry Grass Forest (20) would probably 

have consisted of a diverse range of grasses and 

herbs, with shrubs common only around rocky 

outcrops, as has been described on granite soils of 

the Bathurst area (Semple 1997). Areas of native 

pasture therefore represent remnants of these 

vegetation types and, where grazing pressure and 

other disturbance (such as cultivation and fertiliser 

application) has been light, they may be still quite 

diverse and relatively weed-free. While it is unlikely 

that there were any areas of naturally occurring 

native grasslands in Bega Valley Shire, those areas of 

secondary grassland resulting from the removal of 

the tree layer from grassy open forests which are still 

in good condition represent valuable remnants of the 

original vegetation. 

Loss of biodiversity is likely to have been greatest 

from those vegetation types which occur on the most 

productive sites within the agricultural areas. This 

would mean that 18 (Brogo Wet Vine Forest) and 1 

(Dry Rainforest) are less likely to have been affected 

since they tend to occur on steep and often rocky 

sites. While there are certainly areas of these 

vegetation types which have obviously been cleared 

in the past, they are likely to have suffered lower 

grazing pressure and been allowed to regenerate 

more rapidly. Conversely, riparian vegetation types 

and those associated with moist fertile soils, such as 

19 (Bega Wet Shrub Forest), are likely to have been 

most affected by intense grazing pressure and 

cultivation. 

59


Clearing of riparian vegetation would most likely 

have concentrated on the eucalypt or casuarina forest 

occurring on the river banks. It is unlikely that there 

was a deliberate attempt to remove Riparian Scrub 

from stream beds. However, browsing and trampling 

by livestock with unrestricted access to streams 

would have reduced riparian vegetation. Substantial 

changes to the configuration of streams resulting 

from erosion and sedimentation would have stripped 

or buried riparian vegetation on a large scale. 

Wetland vegetation types have arguably suffered the 

greatest destruction by agricultural activities. Study 

of early portion plans and anecdotal evidence from 

long-term residents (Fryirs 1995; Brierley & Murn 

1995) suggests that smaller streams in the 

agricultural areas consisted of chains of ponds 

distributed along broad flat swampy valley floors 

rather than well defined watercourses. Such areas 

could have amounted to quite a substantial 

proportion of the agricultural areas in aggregate. 

They would have supported a distinctive wetland 

vegetation consisting of a combination of swampy 

meadows dominated by grasses and herbs, reed beds 

dominated by Phragmites australis or Juncus spp, 

sedgelands, and tall scrub dominated by Melaleuca 

ericifolia. Very little of this vegetation remains today. 

Erosion has converted many drainage lines to deeply 

incised channels, often with little vegetation on their 

unstable walls and floors, removing most of the 

habitat for wetland plants. Where gully erosion is 

discontinuous and eroded material has been 

deposited on valley floors, some recolonisation by 

wetland plants has occurred. However, these sites are 

often dominated by introduced species such as 

jointed rush (Juncus articulatus) or pasture grasses 

such as paspalum. Regenerating stands of Melaleuca 

ericifolia are not uncommon in parts of the Wolumla 

catchment, but these are very unlikely to carry any of 

the species found in the understorey of the original 

melaleuca scrubs. Only two mature stands of 

melaleuca scrub with reasonably intact understorey 

are known from agricultural areas (at Jellat Jellat and 

the upper Frogs Hollow Creek swamp on the 

outskirts of Wolumla) and both are very small and 

degraded by cattle trampling and weed invasion. 

Similar vegetation is present in small streams on the 

coastal strip where it may be more extensive, though 

similar pressures (clearing, livestock access, weed 

invasion) occur. 

60 

Altered fire regimes may have contributed, along 

with clearing and grazing pressures, to loss of 

biodiversity in agricultural landscapes. While 

marginal lands have been routinely burnt to provide 

green pick for stock in the past and still are burnt for 

fuel reduction by some landholders, most remnant 

vegetation embedded within agricultural lands has 

probably been fire-free for many years, if not 

decades. This could have caused the loss of some 

species which rely on fire events to stimulate 

flowering or germination of seed. Recent work on 

management of remnant native grasslands indicates 

that some species respond well to regular burning 

and tend to disappear under a fire-free regime. This 

is particularly obvious in areas with railway 

easements which tend to be managed by frequent 

burning and no grazing. Lunt (1995) found that in 

the Bairnsdale area of East Gippsland remnant 

vegetation of grassy forests had tended to segregate 

into a suite of species found on regularly burnt 

railway easements. Another suite was found in 

largely unburnt but lightly grazed remnants. Possibly 

in the absence of railway easements in this area such 

species have already been lost from the local flora, 

but the recent reduction in burning of roadsides in 

the agricultural areas may be a cause for some 

concern in this respect. On the other hand, it has 

permitted tree and shrub regeneration to develop on 

roadsides, where it was formerly probably inhibited 

by burning, as well as routine slashing. 

Another possible cause for concern is loss of genetic 

diversity, not only through clearing and 

fragmentation, which is an obvious cause, but 

through planting of local species grown from 

non-locally-sourced seed. Where remnant vegetation 

is sparse this could have the effect of swamping the 

local gene pool if large numbers of 

non-locally-derived trees are planted. The Farm 

Forestry Project conducted within the agricultural 

areas over the past few years is likely to be the largest 

contributor to this effect, but plantings for 

windbreaks or Landcare plantings could also be 

involved. The Farm Forestry Project has attempted 

to source seed of locally occurring species from areas 

close to the Shire, but this has not always been 

possible. Some seed of local species has come from 

Tasmania (Acacia melanoxylon), as well as from East 

Gippsland, the Southern Tablelands and 

Eurobodalla areas (for example, Eucalyptus


tereticornis, E. viminalis, E. bosistoana), and has been 

planted into areas where those species occur 

naturally (Louise Maud, Farm Forestry Project, pers. 

comm.). With a total planting area of only 100 

hectares in the Shire to date, of which these species 

form only a small proportion, impacts are unlikely 

to be significant at this stage. However, if future 

large-scale farm forestry plantings occur, local 

provenance seed should be obtained. 

The likely impacts of mixing gene pools from 

different areas are not known, but the precautionary 

principle suggests that if a possible negative impact is 

suspected, the activity should be avoided. Closer 

settlement of agricultural areas in recent years, and 

the rise of the Landcare ethos, has resulted in an 

increased amount of tree planting. Some have been 

of local species from more or less local seed 

(plantings done by Landcare groups under 

government grants) but it is likely that the majority 

of plantings by private individuals have been either 

of Australian natives from outside the area or local 

species from non-local seed. The long-term impact 

of this activity on remnant vegetation may be a cause 

for some concern. 

3.1.4 Loss of biodiversity – fauna 

A considerable amount of fauna habitat has been lost 

from agricultural areas through conversion to 

pasture. Lunney and Leary (1988) have documented 

extinctions of mammals associated with European 

settlement of the district (parma wallaby, red-necked 

pademelon and eastern quoll) as well as substantial 

population reductions for a number of species 

(koala, tiger quoll, southern brown bandicoot, 

brush-tailed phascogale and flying fox). However, 

with the exception of the koala (which also inhabits 

woodlands), all of these species are forest dwellers for 

whom substantial areas of potential habitat still exist 

in surrounding public lands. However, the fact that 

European settlement had such a major impact on 

these species suggests that remaining forests in the 

area may not provide optimal habitat for them. 

Management of large areas for timber production 

may affect the habitat values of remaining forests. 

Because the area was originally forest rather than 

grassland or woodland, very few obligate woodland 

or grassland species occur in the area. There are 

therefore no threatened fauna species in the area 

which are solely dependent on agricultural lands for 

habitat (as there are on the tablelands and slopes of 

New South Wales – for example, superb parrot, 

regent honeyeater, eastern earless dragon, striped 

legless lizard). 

A number of species occur which have a preference 

for agricultural areas, and for some of these species, 

land clearing has been advantageous. In fact some 

species, such as the galah, only appeared in the area 

following clearing. Most of these species have no 

difficulty crossing cleared land, so habitat 

fragmentation does not present a problem for them, 

although reduction in the area of grassy forests has 

possibly resulted in population declines for some 

species. Species which have experienced some loss of 

habitat from clearing that cannot be provided in 

surrounding public lands are the diamond firetail, a 

small grass-seed eating finch of grassy woodlands, 

the echidna and the larger macropods, eastern grey 

kangaroo and red-necked wallaby. Fragmentation 

may affect the amount of habitat available to these 

species by inhibiting movements among areas of 

suitable habitat. Another group which may have 

experienced habitat loss is the migratory bird species 

which use the agricultural areas in preference to 

surrounding forests in summer. This group includes 

the pallid cuckoo, Horsfield bronze cuckoo, 

channel-billed cuckoo, dollarbird, white-throated 

gerygone, rufous whistler and tree martin. However, 

although these species depend on remaining forest 

vegetation within agricultural areas, they are able to 

fly over cleared areas, so that fragmentation of 

remaining habitat does not present much 

impediment to their movement. It is important to 

maintain whatever forested connections do occur in 

agricultural areas, to maximise the number of species 

and individuals which can live in and move through 

these areas. 

There are a number of areas within the agricultural 

lands where forest is more abundant, even though 

somewhat fragmented. The most substantial is the 

Brogo area, where forested private land almost 

bridges the gap between Wadbilliga National Park 

and Biamanga National Park/Mumbulla State Forest 

in the vicinity of the Brogo Pass. Similarly the area 

between the former Murrabrine State Forest (now 

National Park) and Murrah State Forest (now 

partially National Park) is almost bridged by 

61


privately owned forested land. Other narrower and 

more tenuous connections occur: 

• between Brogo and Bournda Nature Reserve 

(Black Range) via a series of small remnants on a 

belt of metasedimentary rock which runs 

through Springvale to the west of Bega; and 

• from Moran’s Crossing along the Bemboka River 

and Tantawangalo Creek, both of which carry 

substantial stands of riparian vegetation along 

some sections of their banks. 

These narrower and more fragmented corridors may 

not enable movement of forest-dependent fauna 

from one side of the cleared Bega valley to the other, 

but are still potentially valuable for fauna of more 

open areas, and for migratory birds. 

Clearing of trees from agricultural areas has resulted 

in disruption to the normal succession of tree age 

classes which occurs in forests. While there are still 

substantial numbers of old trees and dead trees 

carrying hollows to provide nesting and roosting 

sites for fauna, there are few mature trees present to 

replace them. There are considerable numbers of 

young trees and saplings now present in many areas, 

but it will be well over 100 years before these begin 

to form hollows. In that time many more of the 

existing hollow-bearing trees could have died and 

fallen or been cut down, so that there is likely to be a 

period when there is a shortage of hollows available 

to the fauna which require them. This fauna includes 

all parrots and cockatoos, kookaburra, sacred 

kingfisher, dollarbird, owls, owlet nightjar, wood 

duck, striated pardalote, tree martin, white-throated 

treecreeper, most insectivorous bats, common 

brushtail possum and sugar glider. Many of these 

species perform a useful function in consuming 

insects on farms. As a result, population declines 

resulting from a loss of breeding habitat may have 

consequences for farm productivity, as well as for the 

continuing health of remaining remnant vegetation. 

It is therefore important to manage remaining old 

and dead trees, to maintain them for as long as 

possible. 

Grazing within remnant forest has caused the loss of 

feeding and nesting habitat for many fauna species 

due to: 

62 

• changes to understorey structure with loss of 

shrubs from areas where they formerly occurred; 

• removal of woody debris; and 

• reduction of leaf litter on the ground. 

Loss of understorey also removes habitat for species 

which feed on damaging insect pests, exacerbating 

dieback and failure of regeneration in agricultural 

landscapes. 

As freshwater wetlands have been much reduced in 

the area, fauna dependent on these habitats are likely 

to have substantially declined. Information is not 

available, but one might assume that some frog 

species have declined, apart from the threatened 

green and golden bell frog (Litoria aurea), whose 

population crash throughout its range in the early 

1980s has been well documented. Birds which 

frequent reed beds (golden-headed cisticola, little 

grassbird and reed warbler) still occur in the district, 

but the habitat available to them has been drastically 

reduced. Other aquatic fauna (such as various 

freshwater fish species, platypus and water-rat) have 

obviously suffered similar losses of habitat. The lack 

of structural diversity within streams, resulting from 

sedimentation and loss of woody debris from river 

systems as a result of clearing in their catchments, 

has severely reduced feeding and breeding resources 

available for aquatic fauna. 

3.1.5 Weed invasion 

An environmental weed is a weed which invades 

native vegetation, as opposed to those which inhabit 

agricultural areas. This very wide definition means 

that many agricultural weeds, and even useful 

pasture species, may also be regarded as 

environmental weeds where they are growing in 

remnant native vegetation rather than in improved 

pasture or crops. While most environmental weeds 

are introduced species, there is considerable scope 

for native Australian plants to become 

environmental weeds in areas outside their natural 

range. This is because many environmental weeds 

started out as garden escapees, and there has been an 

increasing trend in recent years to gardening with 

natives, and to planting trees on rural properties and 

roadsides.


In Bega Valley Shire environmental weeds are most 

obvious around towns and villages, where long 

settlement has provided a source of seed from 

gardens. This is particularly true where creeks and 

rivers flow through towns, since the more fertile soils 

and more reliable soil moisture near watercourses 

encourages the establishment of introduced species. 

Examples can be found on the riverbanks in Bega, 

Candelo, Cobargo and Quaama, where remnant 

native vegetation has been invaded (as in Quaama) 

or largely replaced (as in the other three towns) by a 

suite of exotic species. 

One of the reasons why environmental weeds are a 

problem is their ability to spread beyond the area in 

which they were originally planted. Many have 

fleshy fruits which are eaten by birds or mammals 

such as foxes, and then distributed in the animal’s 

droppings. Examples are privet, cotoneaster and 

blackberry. Others have seeds which are carried long 

distances by wind (for example, tree of heaven, moth 

plant, Cape ivy) or are spread by vegetative means – 

either through pieces of plant adhering to passing 

animals (for example, tiger pear) or by being carried 

in flood waters (for example, wandering jew, willow, 

periwinkle, common prickly pear). 

The impact which environmental weeds have on 

native vegetation depends on the level of infestation. 

At low levels of abundance the impact may be 

minimal, but at higher densities weeds displace 

native plants which could have grown on the site. 

This may so alter the characteristics of the site (for 

example, by casting dense shade) that it is impossible 

for native plants to survive there. Exotic vines, by 

climbing over native trees and shrubs, can so shade 

their foliage that the plants are no longer able to 

photosynthesise and may die. 

Compared to urban areas Bega Valley Shire is still 

relatively lightly infested with environmental weeds 

of bushland, partly because of low population levels 

and low levels of disposable income spent on 

gardening. However, this is changing rapidly. It can 

be expected that the existing environmental weeds in 

the district will go on expanding their ranges, and 

new weeds will be introduced via gardens and 

plantings on rural lots and roadsides. The district has 

few natural barriers to the spread of such weeds, 

having a relatively high rainfall and relatively fertile 

soils. Many environmental weeds are incapable of 

surviving in infertile soils (such as those derived 

from the Sydney sandstones) unless nutrient levels 

are artificially raised by run-off into bushland. This 

inhibition on their spread is less pronounced in the 

Bega valley area. However, weed invasion still tends 

to be much worse in riparian situations or in 

drainage lines where soil moisture and nutrient 

availability are less limiting, or in residential areas 

where nutrient-enriched run-off enters native 

vegetation. Riddell (1997) has documented the 

particular threat which willows pose to riparian 

communities in the region. That report includes 

maps of the occurrence of mature and seedling 

willows in all major streams of the Shire. 

Some plants which are serious weeds in other 

districts have not yet shown the same degree of 

invasiveness in this area, and it may be that the 

climate here is less suitable for them. Many weeds of 

the drier tablelands and slopes are found here only in 

the driest part of the Bega valley, around Candelo 

(for example, gorse, pepper tree, African boxthorn). 

Others are abundant to the north, but as yet occur 

only sparsely or are localised at a single infestation 

site in Bega Valley Shire (for example, crofton weed, 

lantana, wild tobacco bush, arum lily). 

As yet, Bega Valley Shire has few of the serious water 

weeds. However, because water weed seed or 

fragments can be spread in mud on the feet of 

waterbirds, it is realistic to assume that sooner or 

later at least some of them will arrive in the district. 

The recent discovery of sagittaria (Sagittaria 

graminea ssp. platyphylla) on the shores of Brogo 

Dam is an example of this. Salvinia (Salvinia 

molesta), which has been a major problem on 

remnant wetlands in Eurobodalla Shire, has been 

found on a dam at Bunga Head. A water hyacinth 

(Eichhornia crassipes) infestation on Glebe Lagoon in 

Bega was fortunately eradicated before floods could 

spread it into the Bega River, though not before 

demonstrating considerable potential to spread on 

the lagoon (Alan Smith, BVSC, pers. comm.). 

Disposal of aquarium plants into dams and natural 

water bodies is another possible source of such 

infestations. 

Native grasslands are particularly vulnerable to weed 

invasion, since most agricultural weeds can survive 

63


in them, although many need some sort of 

disturbance in order to gain a foothold. However 

disturbance is seldom absent, since almost all native 

pasture is grazed. While native grasslands are 

unlikely to have occurred naturally in the Bega 

valley, some parts of the district show a high diversity 

of native grasses and herbs. Although these may be 

secondary grasslands or woodlands created by 

clearing of the original grassy open forests, they are 

still worthy of preservation where they occur, given 

the poor conservation status of native grasslands in 

temperate Australia. Invasion by aggressive exotic 

grasses such as African lovegrass, serrated tussock, 

Yorkshire fog and kikuyu threatens the integrity of 

remaining stands throughout the agricultural areas. 

The same applies to remnant forest with a grassy 

ground cover. The need to combat invasive exotic 

grasses with herbicides may cause the loss of native 

pastures which have managed to persist up to the 

present under light grazing regimes. Grassy 

communities of drainage lines are particularly 

susceptible to weed invasion and are unlikely to be 

in good condition anywhere in the agricultural areas. 

Appendix 4 (p. 125) lists species currently known to 

occur in Bega Valley Shire which are, or show the 

potential to become, environmental weeds. 

Numerous agricultural weeds which invade remnant 

vegetation have been omitted from this table if they 

are generally too small to seriously exclude native 

species or if they only penetrate remnant vegetation 

in the presence of considerable disturbance. 

3.1.6 Impact of nutrient-laden 

run-off and sedimentation on 

wetland vegetation 

Many of the coastal lakes and estuaries have 

residential development within their catchments 

and, since most are low discharge drainage systems, 

they are frequently closed to the sea by sand bars. 

The potential for nutrients to accumulate is 

therefore high, and it seems probable that this will 

have adverse effects on the aquatic vegetation (sea 

grasses) of these systems, and possibly also on 

fringing vegetation such as mangroves and salt 

marsh. As yet, few such impacts have been recorded 

in the Shire, but with increasing population it seems 

probable that they will occur, either through 

discharge from sewage treatment systems (like the 

64 

Tathra system) or through seepage from septic 

systems in unsewered areas (such as at Wallaga 

Lake). Patterson, Britton and Partners (1996) report 

that nutrient loads (nitrogen and phosphorus) 

frequently exceeded ANZECC guidelines in Mead’s 

Bay, although levels in the remainder of Wallaga 

Lake were low. Excessive growth of epiphytic algae 

on sea grasses, and sparser populations of sea grasses 

(possibly as a result of algal smothering), were noted 

in Mead’s Bay. 

Livestock access to wetland vegetation in coastal 

areas is another problem mentioned in the Wallaga 

Lake report. Eating and trampling of wetland 

vegetation, stirring up sediments and introducing 

nutrients could all have substantial impacts. Several 

Shire wetlands covered by State Environmental 

Planning Policy 14 appear to be subject to grazing, 

which is permitted under existing use rights. 

The encroachment of sand into coastal lakes and 

estuaries as a result of clearing and subsequent 

erosion and sedimentation in their catchments may 

cause loss of habitat for seagrass communities. 

Accelerated rates of fluvial delta extension into the 

lakes from in-flowing streams have been reported for 

Wallaga Lake (Patterson, Britton & Partners 1996) 

and Pambula Lake (Thoms & Bergs 1994). 

However, the influx of sediment into the Bega River 

estuary, which also supports seagrasses, is probably 

more significant given the very large bed-load of 

sand present in this system. Estuary processes in the 

Bega River have not yet been investigated. 

3.2 Currently forested 

landscapes 

Threats to vegetation in areas which have not been 

cleared for agriculture include: 

• clearing (for urban or rural residential 

development, or for agriculture); 

• timber harvesting activities and associated 

manipulation of fire regimes; 

• dieback of forest trees; 

• invasion of weeds from more disturbed land 

nearby; and 

• increased nutrient inputs into adjacent coastal 

aquatic ecosystems.


3.2.1 Clearing 

While some limited clearing for agriculture is still 

occurring in parts of the region, probably the 

greatest source of pressure to clear vegetation comes 

from developers wishing to create rural residential 

subdivisions, particularly within the coastal strip. 

Such developments usually result in the removal of 

all understorey and a large proportion of the trees. 

Even where lot size is relatively large and some bush 

is initially left undisturbed, subsequent clearing after 

sale is likely to be extensive. Under Bega Valley 

Shire’s guidelines for building in bushfire-prone 

areas, clearing a 40-metre radius around a house site 

is required, with a further 60 metres to be 

maintained free of fuel (that is, understorey). In 

addition to this mandatory clearing, there is likely to 

be clearing along fence lines, service easements and 

access tracks and for dams, and possibly some 

creation of pasture for livestock. 

Because of pressure to retain viable agricultural land 

in agricultural production, zoning for rural 

residential development (1(c) zone) within Bega 

Valley Shire has been concentrated within forested 

land on or close to the coastal strip. The vegetation 

types most likely to be represented within these areas 

are: 

• 32, Coastal Foothills Dry Shrub Forest (five per 

cent cleared, 24 per cent reserved); 

• 34, Coastal Gully Shrub Forest (12 per cent 

cleared, 25 per cent reserved); 

• 37, Lowland Gully Shrub Forest (six per cent 


• 46B, Lowland Dry Shrub Forest (five per cent 


• 47, Eden Dry Shrub Forest (four per cent 


• 49, Coastal Dry Shrub Forest (two per cent 


• 36, Dune Dry Shrub Forest (41 per cent cleared, 

23 per cent reserved, but this is an overestimate); 

• 13, Hinterland Wet Fern Forest (nine per cent 


• 14, Hinterland Wet Shrub Forest (four per cent 

cleared, 39 per cent reserved); and 

• 6, Coastal Warm Temperate Rainforest (one per 

cent cleared, 42 per cent reserved). 

Of these, 32 and 34 are more common in the 

northern half of the Shire and 37, 47 and 49 are 

more common in the southern half. The others 

occur throughout. The wetter types (13, 14 and 6) 

are most likely to occur in sheltered gully sites and, 

as such, may be protected from clearing to some 

extent by topographic site constraints. However, 

they will still be subject to flow-on effects from 

nearby clearing and settlement, such as altered fire 

regimes, weed invasion, and increased tree mortality 

from exposure to weather extremes, and possibly 

increased attractiveness to bell miners (see 

section 3.2.4 on dieback). 

It can be seen from the above figures for clearing and 

reservation that most of these vegetation types are 

thought to have been only slightly affected by 

clearing up to this time. The exception is 36, Dune 

Dry Shrub Forest – a naturally restricted type to 

begin with – of which some 40 per cent is thought 

to have been cleared. However, the mapping for this 

vegetation type was found to be rather inaccurate 

due to problems with the geological layer used. It 

appears that extant stands may in fact be 

considerably less extensive than thought. The 

reservation status of this type is therefore 

considerably worse than suggested by Keith and 

Bedward (1998). It is the most vulnerable coastal 

vegetation type, with most stands occurring between 

Bermagui and Cuttagee Lake in a narrow strip. At 

least some of this area is vacant crown land, so there 

is some scope for further reservation of this 

vegetation type. Some clearing of this type has 

occurred for urban development in Merimbula. The 

only other sizeable areas of 36 are on private 

property at Wallagoot Lake, and between the Nullica 

River mouth and Boydtown. Prevention of further 

clearing of this vegetation type should be a priority. 

Of the widespread coastal vegetation types (32, 34, 

37, 46B, 47 and 49), all are thought to occur over 

fairly large areas (14 000 to about 32 000 hectares 

extant), to be little cleared and to meet the 

reservation criterion of 15 per cent of their pre-1750 

extent in reserves. There seems at first sight to be 

65


little cause for concern if clearing of these vegetation 

types occurs for residential developments. However, 

this does not take into account their habitat 

potential for fauna, and particularly for threatened 

species. 

The coastal strip is a major migration route for 

migratory birds, particularly honeyeaters, which 

move through the area in very large numbers in 

spring, autumn and winter. While flowering 

banksias provide a food resource, flowering eucalypts 

are also important. There appears to be a 

concentration of autumn–winter–spring flowering 

species (spotted gum, woollybutt, ironbarks and red 

bloodwood) in vegetation types 32, 34 and 46B. 

This makes these types particularly significant for 

migratory birds (not only honeyeaters, but also 

insectivorous species which feed on insects attracted 

to the flowering trees). These types are also 

important for mammals which feed on nectar, 

including eastern pygmy possum, feathertail glider 

and the threatened species, yellow-bellied glider. The 

understorey tree Allocasuarina littoralis, the seeds of 

which are the principal food for the threatened 

glossy black cockatoo, also tends to be concentrated 

within the coastal strip in vegetation types 32, 37, 

46B and 49. 

Constraints to development within these coastal 

vegetation types thus tend to revolve around fauna 

rather than vegetation issues. These issues include 

the need to demonstrate a lack of significant impact 

of proposed developments on populations of 

threatened fauna species and the need to maintain 

forested links (‘wildlife corridors’) through 

developed areas to maintain the possibility for 

genetic exchange and movement of individuals 

between populations by the resident fauna. 

Areas along the coastal strip which appear (from 

perusal of maps) to be potentially important for 

wildlife corridors are: 

• vacant crown land between Pambula and 

Merimbula, and north of Merimbula; 

• the Bald Hills area west of Pambula and an area 

west of South Pambula; 

• an area of vacant crown land and private 

property straddling the Princes Highway 

between Wolumla and Merimbula; 

66 

• the area of private land around the western side 

of Wallagoot Lake and between Bournda Nature 

Reserve and the northern section of Bournda 


• private property and vacant crown land between 

Kalaru and Tathra, through to the Bega River; 

and 

• forested private property around Rileys Road 

between Cobargo and Bermagui which links the 

northern end of the former Murrah State Forest 

(now National Park) and Wallaga Lake National 

Park. 

There are areas away from the coastal strip where the 

vegetation consists largely of forest and the tenure is 

mostly private property. Many of these areas are also 

interspersed with State Forests and reserves in such a 

way as to make them potential wildlife corridors. 

However, being in freehold tenure makes them 

vulnerable to clearing, possibly fragmenting or 

severing these linkages. These areas include: 

• lower parts of the Towamba valley between 

Towamba and Kiah; 

• upper parts of the Towamba valley between New 

Buildings and Burragate; 

• the area along Burragate Road between 

Burragate and Wyndham; and 

• the Yuranglo Voluntary Conservation Agreement 

area at Devil’s Hole south of Myrtle Mountain. 

Some of these areas also include vegetation types of 

higher conservation significance, most notably 19, 

Bega Wet Shrub Forest, which is mapped in all these 

areas, but is most common along the Burragate 

Road. 

As in agricultural areas, the Native Vegetation 

Conservation Act 1997 provides some constraints on 

clearing, which are discussed in section 4.1 below. 

3.2.2 Timber harvesting 

Timber harvesting occurs principally within State 

Forests, although it has also often occurred as an 

adjunct to clearing for subdivision on private lands. 

With increasing interest in the district in farm 

forestry, and loss of public production forest into the 

reserve system, it may become a more common land


use on private land. The need to satisfy the 

requirements of the Native Vegetation Conservation 

Act 1997 may be perceived as a disincentive, though 

it does not apply to selective logging. 

Forestry practices in public forests in the Eden 

Management Area have been contentious since the 

advent of wood chipping in 1969. Prior to this time 

logging had been selective and limited in area by 

accessibility. The advent of a market for poorer 

quality trees prompted more intensive felling 

practices and provided the stimulus for roading into 

previously untouched areas. Most criticism of recent 

forestry practice has concentrated on the impact it 

might have on fauna populations, on streams 

(through erosion and sedimentation) and on water 

yields from catchments carrying very dense stands of 

young regrowth. There has been less discussion of 

direct impacts on the vegetation itself. 

There is an assumption within the forest industry 

that the forest which regrows after logging is 

virtually identical to that which was logged. There 

have been some studies to test this, though only one 

has been reported on by State Forests of New South 

Wales in this area (Bridges 1983). This study 

concentrated on the various silvertop–stringybark 

forest types dominated by Eucalyptus sieberi, 

E. globoidea and E. agglomerata. The conclusion 

derived from inspection of regeneration in 7000 

hectares of logged forest was that the species 

composition of the regeneration is very similar to 

that of the original stand. However, this refers only 

to tree species; there is no discussion of possible 

impacts on understorey. There has apparently been 

research conducted in the Eden Management Area 

on logging impacts on species composition, 

commencing in 1977, but the data has not yet been 

analysed (D Binns, SFNSW, pers. comm.). 

Another study conducted in the same forest type in 

Timbillica and Yambulla State Forests (de Chazal 

1992) found no differences in tree species 

composition between areas which had been burnt by 

wildfire and logged areas, and only minor (not 

statistically significant) differences were found for 

understorey. However, de Chazal points out that this 

is not conclusive, since changes could have been 

masked by the more obvious differences in 

understorey composition between sites due to other 

factors such as geology and site exposure. It was not 

possible to find sites which were identical in all 

respects other than logging or fire history. 

Studies from outside the Eden Management Area 

tend to confirm that logging impacts on 

silvertop–stringybark forest types are minor, since 

these forest types are adapted to recover from 

frequent wildfires (for example, Loyn et al. 1983 in 

East Gippsland). However some studies show an 

effect on species composition, at least in the short 

term, for other forest types. Intensive logging 

practices have not been in operation for long enough 

to provide information about long-term impacts. 

Mueck and Peacock (1992) compared similar sites of 

different ages since logging (2, 5, 10, 18–22 and 

25–29 years) with unlogged sites within the same 

vegetation type, for three different forest types in 

East Gippsland. The forest types were referred to as: 

• Lowland Sclerophyll Forest (similar to local 

silvertop–stringybark types); 

• Damp Sclerophyll Forest (not directly 

comparable to any single local vegetation type); 

and 

• Wet Sclerophyll Forest (similar to local 

escarpment forest types, particularly 10, 

Mountain Wet Layered Forest). 

They found no significant effect of logging on 

species composition within Lowland Sclerophyll 

Forest, but significant impacts on species 

composition for the other two forest types. Both 

types tended to shift in composition towards drier 

forest types. This effect reduced over time, but had 

not disappeared at the 30-year time limit of the 

study. Eucalyptus sieberi appeared in two-thirds of 

the regrowth plots in Damp Sclerophyll Forest, 

though it was in none of the control plots. It also 

appeared in Wet Sclerophyll Forest regrowth plots. 

The authors conclude that under the current 

harvesting rotation length of 80–100 years, with 

probable interim disturbance at 40 years for 

thinning, original species composition will be 

unlikely to re-establish in some forest types. 

Ough and Ross (1992) investigated impacts of 

clear-felling and subsequent burning of coupes in 

wet montane forests of the Victorian Central 

Highlands, also comparing sites with varying age 

67


classes of trees, from one year to an estimated 250 

years. Most vegetation in the study fell into the 

category of Wet Sclerophyll Forest. Findings were 

that most species present prior to logging were also 

present in the regeneration, though proportions 

changed significantly in some cases. However, a 

cause for concern was the impact of logging on tree 

ferns (Dicksonia antarctica), which are killed by 

mechanical damage during logging, are incapable of 

resprouting from the base, and are very slow to 

return to the site from spores, and to grow to their 

original size. Ough and Ross conclude that 

clear-felling rotations of 50–80 years may jeopardise 

the predominance of this species, which is very 

long-lived, and often dominant in the understorey of 

wet sclerophyll forests. This could have implications 

for the presence of other species in regrowth wet 

sclerophyll forest, since many epiphytic ferns grow 

only or largely on tree fern trunks. Many shrub and 

tree species germinate on tree fern trunks, including 

Cool Temperate Rainforest species, so that changes 

to tree fern abundance may affect the development 

of rainforest on sheltered sites within these forests. 

Similar forests are present on the escarpment within 

the Eden region, so these findings are applicable to 

this area. Although these forests are now reserved 

within the South East Forests National Park, they 

were logged prior to the change in tenure. 

A report on regeneration in moist tableland forest 

types (FCNSW 1983) states that many of the tree 

species found in these forests do not carry regular 

seed crops, and that in some years there may be 

insufficient seed in the canopy to ensure adequate 

regeneration after logging. However, it does not 

discuss the impact this might have on species 

composition. 

In general, the literature tends to suggest that the 

silvertop–stringybark forests which form a 

substantial proportion of the publicly owned 

production forests in the Eden region are quite 

resilient to intensive logging, at least in terms of their 

species composition. However, there may be 

significant impacts from logging on species 

composition in other forest types which are also 

widespread in the Eden region. Privately owned 

forests could be expected to fall more into the forest 

types whose composition may be changed by 

logging, since the silvertop–stringybark forests occur 

68 

mostly on the poorer soils which have tended to 

remain in public ownership. Changes to plant 

species composition may also have effects on fauna, 

but there has been little or no research on this 

subject. The probable impacts on fauna of changes 

to forest structure, with loss of tree hollows and 

replacement of multi-aged forests with uniform 

stands of young trees, has received more attention. 

Forestry practices have been amended somewhat as a 

result, with substantial reductions in coupe size since 

the early clear-felling practices and more retention of 

unlogged strips along drainage lines and of ‘habitat 

trees’ within coupes. However, whether these 

prescriptions are adequate to maintain populations 

of forest-dependent fauna in production forests in 

the long term is not known (Gibbons 1994; Recher 

et al. 1987; Gibbons & Lindenmayer 1997). The 

Eden Region Forest Agreement (Anon 1999) 

requires that there be ongoing monitoring, with 

reporting at five-year intervals, of the sustainability 

of management practices in production forests. This 

is to include monitoring of threatened fauna species 

populations and the extent of connectivity of 

retained habitat for these species. 

3.2.3 Fire regimes 

Fire regimes (frequency, intensity and seasonality) 

play a key role in determining the structure, floristics 

and extent of vegetation associations in the region. 

Vegetation associations are all adapted to the 

particular fire regimes which have prevailed in the 

past. Fire regimes have changed since European 

settlement and hence we can expect some level of 

flora and fauna change, including decline in species 

diversity or even extinction of species where the 

effects are widespread over long periods of time. 

The community expects a level of protection from 

potential damage due to wildfires. Fuel reduction 

activities, including hazard reduction and other 

techniques, are carried out within this area with the 

objective of reducing the risk to life and property, or 

maintaining the economic timber resource in 

production forests. 

Prescribed use of fire should also achieve ecological 

objectives. There have been a number of studies of 

the impacts of fire, and of fire regimes on particular 

plant species and plant associations. However, to 

design and then maintain particular management


fire regimes to maintain biodiversity in the range of 

vegetation associations present in the region is a 

complex task which is currently being approached by 

fire management agencies. A significant impediment 

to the design of such programs is the current lack of 

research on the fire regimes required for the range of 

vegetation types in the Bega valley region. 

It is understood that the regular use of fire to 

maintain reduced levels of fuel will, in time, reduce 

species diversity in some parts of the region, as will 

the absence of fire over long periods for other parts. 

It is clear that firefighting agencies must work 

together to identify appropriate fire regimes for each 

vegetation association, threatened flora or fauna 

species. 

Under the Rural Fires Act 1996, fire management 

agencies are expected to work in a coordinated 

fashion to plan for and then implement fire 

management programs on a cooperative basis to 

achieve protection of life and property whilst still 

maintaining natural processes. Locally, rural fire 

services (including brigade members) and staff from 

the Department of State Forests and the National 

Parks and Wildlife Service work together to achieve 

joint programs whilst still achieving the corporate 

aims of each individual agency. 

In the Bega Valley Shire controlled fire is used 

routinely by landowners, brigades and fire 

management agencies for fuel reduction around 

towns and rural properties, to reduce the risk of 

property damage in the event of a wildfire. The Bega 

Valley District Fire Committee has a clear objective 

to establish long-term strategies to enable the 

continued use of fuel reduction to achieve life and 

property protection whilst still protecting species 

diversity. 

Little is known about the possible effects of existing 

fire regimes on forest ecosystems. However, from the 

extensive literature on the impacts of frequent fire on 

individual species (reviewed in Keith 1996), an 

inference could be drawn that localised extinctions 

of some species (and hence shifts in understorey 

composition) are likely. 

State Forests of New South Wales has been 

conducting research into the impacts of logging and 

fire regimes on understorey plants in the Eden 

region since 1986. Trial plots have been burnt at 

nominal two-year and four-year intervals and the 

effects on the vegetation, and on individual plants, 

have been monitored. The trials are being conducted 

in silvertop–stringybark dry forest. This research is 

continuing. 

While fire regimes which are intended primarily for 

property protection may have adverse ecological 

effects on forests in general, some of the most 

vulnerable vegetation types are those which are 

confined to small areas embedded within larger 

forest tracts. There are two key examples in the local 

area of plant associations which may be significantly 

affected by the application of an inappropriate fire 

regime. 

Rainforest is generally found in small linear strips 

within sheltered gullies or on sheltered slopes at the 

top of the escarpment. Rainforest stands are of 

particular scientific and ecological value on the far 

south coast because many of the species contained 

within them, and some of the rainforest types, are at, 

or close to, their southern limits in this area. Many 

rainforest species have a very limited capacity to 

recover from fire. Damage from wildfires is obvious 

in most rainforest in the district (Floyd 1982). Floyd 

states very strongly that another wildfire could 

destroy many already damaged rainforest stands in 

the district and that fire needs to be managed in the 

district to prevent this. 

Wet heaths are another vulnerable vegetation 

community which may be interspersed in small 

stands through eucalypt forests. In particular, 

vegetation types 56 (Hinterland Heath) and 57 

(Lowland Swamp) are found mostly in forested areas 

south-west of Eden. Because of the abundance of 

fine fuels, which are well aerated and uniformly 

distributed from ground level to the canopy, heaths 

are extremely flammable. Consequently, cool burns 

in adjacent forest could become much more intense 

in heaths. While heathland species are generally well 

adapted to fire, too-frequent fires could result in 

local extinction of some species. 

Other vegetation values in the local area may also be 

compromised by the current fire management 

regimes, including loss of middle-storey species, old 

growth values or other habitat features such as tree 

hollows. 

69


3.2.4 Dieback 

The phenomenon of dieback occurring in forested 

areas, rather than among remnant vegetation of 

agricultural areas, has been recognised within the 

district in recent years, and has been attributed to 

bell miners (Manorina melanophrys). These native 

birds live in large colonies which defend territories 

against all other birds. Their favoured habitat is 

sheltered gullies containing forest with a tall dense 

shrub understorey, in which they nest. They feed 

principally on lerps (the sugary covering of 

sap-sucking insects called psyllids), on the psyllids 

themselves and on other insects. By driving other 

insectivorous birds out of their territory, and 

possibly by consuming invertebrate predators of 

psyllids (Stone 1996), they permit populations of 

damaging insects to build up to levels where tree 

health is affected. 

It is not known to what extent human activities in 

forests have influenced what appears to be a recent 

increase in bell miner colony numbers. There has 

been speculation (Backhouse 1987) that clearing of 

adjacent areas may place remaining trees under 

stress, encouraging psyllid populations to build up 

(possibly due to increased nutrient levels in the 

foliage of stressed trees), which then allows bell 

miner populations to also increase. Another 

possibility is that reductions in burning frequency in 

recent years have allowed a dense understorey of 

mesophyll shrubs to develop in areas from which 

burning formerly excluded them, favouring bell 

miners. Anecdotal evidence from the north and 

central New South Wales coast, where SFNSW 

research has been conducted to date, suggests that 

partial clearing and creation of permanent water 

sources (for example, farm dams) increases the 

establishment and persistence of bell miner colonies 

(Stone 1999). 

It is also not known to what extent the phenomenon 

constitutes a problem, though in particular 

situations it may be an obvious problem if bell 

miners reduce habitat values for other less common 

fauna. An example is the threatened helmeted 

honeyeater in the Dandenong Ranges, whose 

preferred habitat coincides with that of bell miners. 

One colony of this species has been driven to 

extinction apparently by competition from bell 

miners (Backhouse 1987). Closer to home, it has 

70 

been suggested that koala populations in Murrah 

and Bermagui State Forests may be disadvantaged by 

the presence of numerous bell miner colonies, since 

dieback among preferred food trees reduces the 

available food resource for them (Bertram & Allen 

1998). It is also likely to reduce the timber yield 

from production forests. Observation suggests that 

on the north and central coast rapid changes in 

forest species composition are occurring on affected 

sites, with susceptible eucalypt species being replaced 

by rainforest trees or less susceptible hardwoods such 

as brush box and tallowwood (C Stone, SFNSW, 

pers. comm.) 

It is anticipated that SFNSW will be conducting 

some small-scale research on bell miners in the 

district in the near future under the terms of a 

Natural Heritage Trust grant made to the Bega 

Environment Network Centre to assess all forms of 

dieback in the area. This will involve experimental 

removal of understorey on a single site and 

observation of the effects on tree health (C Stone, 

SFNSW, pers. comm.). From personal observation, 

bell miner-mediated dieback appears common in 

privately owned forests around the margins of 

agricultural areas and on the coast in Bega Valley 

Shire, and also occurs in public lands, although the 

extent has not been surveyed. It appears that partial 

clearing for residential development may 

compromise the health of remaining adjacent forest, 

as may logging operations in some areas. The 

viability of regenerating forest in marginal 

agricultural areas may also be affected by increased 

bell miner activity. 

There is a need for an immediate survey, across all 

tenures, of bell miner colonies and dieback to 

provide baseline data against which future changes 

can be assessed. Under the Eden Region Forest 

Agreement, both SFNSW and NPWS are required 

to review the sustainability of forest management 

practices over time. Sustainability indicators have 

been devised, and must be monitored, providing the 

framework of a five-yearly review process. These 

indicators include forest ecosystem health and 

vitality. The Regional Forest Agreement document 

(Anon 1999) states: 

‘Processes and agents that may change ecosystem 

functioning need to be identified on a regional 

basis. These include interactions between natural


events and management actions in the following 

areas: …logging, clearing, roading, bell miner 

dieback, insects and disease.’ 

There may be some justification for culling bell 

miner populations in specific locations where the 

vegetation being affected by them is of particular 

conservation significance. Examples of such sites 

include a dune swamp with forest red gum 

(E. tereticornis) behind Pambula Beach, and remnant 

Floodplain Wetland vegetation comprising swamp 

gum (E. ovata) with a Melaleuca ericifolia 

understorey immediately south of Kalaru. Similar 

vegetation is also affected on Sapphire Coast Drive 

north of Tura Beach. The latter two sites are on 

private property, while the former is partially in 

National Park and partially in vacant crown land. A 

survey might indicate other areas where vegetation 

of high conservation significance is being affected. 

3.2.5 Weed invasion 

Weed invasion of undisturbed forest is generally 

minimal but, where forest is disturbed or is in close 

proximity to an abundant source of weed seeds, 

weed establishment may occur. Logging within large 

expanses of forest is generally followed by the 

introduction of a small number of relatively 

ubiquitous weeds such as cat’s ear, fleabane and 

thistles (Hypochaeris radicata, Conyza albida and 

Cirsium vulgare). However, where logging occurs 

around forest edges, as on private property, weed 

invasion could be much worse. Weeds such as those 

mentioned above tend to persist only along road 

verges and in creeks within the forest. The 

cumulative effects of continual small introductions 

from logging within the forests of the district is not 

known. 

Instances of weeds invading adjacent forest from 

abundant seed sources nearby occur around coastal 

towns situated among forest (such as Merimbula), 

where exotic species are increasing in abundance in 

adjacent forest. This problem can only be expected 

to get worse, and the district does not have a large 

population to draw on for bush regeneration groups 

to help control the weeds. Design features of 

residential subdivisions can contribute to weed 

invasion problems by: 

• fragmenting forest, thereby increasing disturbed 

edges; 

• channelling increased flows of nutrient-enriched 

run-off into native vegetation; and 

• backing house lots onto forest, making it easier 

to dump garden refuse in the forest. 

Another example of weed invasion of forest from 

adjacent disturbed land is that of serrated tussock 

moving into State Forest and National Park around 

parts of the Towamba valley. 

3.2.6 Phytophthora 

Phytophthora cinnamomi (cinnamon fungus) is a 

fungus which lives in the soil and attacks plant roots, 

reducing their ability to absorb water and nutrients, 

and sometimes causing dieback. Plants vary in their 

susceptibility to the fungus. Phytophthora is 

presumed to be responsible for the phenomenon of 

Jarrah (Eucalyptus marginata) dieback in Western 

Australia, but in general, at least in eastern Australia, 

eucalypts tend to be quite resistant. Understorey 

species, including members of the families Fabaceae 

(peas), Proteaceae (banksias, grevilleas and so on), 

Epacridaceae (heaths) and Xanthorrhoeaceae (grass 

trees) are most susceptible (Hussey & Wallace 1993). 

Susceptibility of individual plants may be increased 

by the operation of other factors causing stress, such 

as drought or waterlogging. 

The fungus is often spread from site to site in soil 

adhering to machinery or vehicles (or even shoes) 

and, once established, is spread outwards from the 

source by the movement of soil water, or by passing 

from one root system to another. Because of the role 

of water in its spread, the resultant dieback can often 

be seen to spread down a drainage line, though it can 

also move more slowly up-slope. 

Information about the occurrence and effects of 

Phytophthora in Bega Valley Shire has been provided 

by K McDougall (NPWS Southern Zone, pers. 

comm.). The fungus has been detected in a number 

of State Forests in the region during sampling by 

SFNSW in the 1970s, but no effect on vegetation on 

the sites was observed and the possible long-term 

effects remain unknown. Only two locations are 

known where plant deaths have occurred and are 

likely to have been a result of Phytophthora infection. 

One occurrence is at Bittangabee and Green Cape in 

Ben Boyd National Park. Sites are in forest adjacent 

to roads, and records of the presence of the fungus 

71


go back to the 1970s. Dead banksias have been 

observed and the fungus has been cultured from the 

roots of a pea species (Pultenaea daphnoides). The 

proximity of this site to extensive areas of heath, 

which are bisected by a road which receives 

moderate levels of vehicle use, especially in summer, 

may be a cause for concern, since heaths contain 

many species from the most susceptible plant 

families. 

The second occurrence is on the summit of Mount 

Imlay, where the fungus was detected in 1995 in soil 

at the end point of a former vehicle access track 

(now overgrown) and from the base of a tree 

72 

showing dieback symptoms (Graham-Higgs & 

Associates 1995). Dead grass trees have subsequently 

been observed in an area down-slope from the 

summit, and further dieback has occurred in five to 

10 per cent of individuals of the rare Mount Imlay 

mallee (Eucalyptus imlayensis). The dieback appears 

to be spreading down a gully from the summit. 

Although the fungus was detected as a result of an 

impact assessment for the Seaphone installation on 

Mount Imlay, it appears that it was present on the 

site prior to the commencement of that 

development. NPWS hopes to undertake fungicide 

trials on both areas in the spring of 1999.


4 Protection mechanisms 

for vegetation 

Various State and local government policies and 

legislation are in place which provide some 

protection for native vegetation. These are outlined 

below. 

4.1 Native Vegetation 

Conservation Act 1997 

The Native Vegetation Conservation Act 1997 (NVC 

Act) is administered by the Department of Land and 

Water Conservation and requires landholders to 

obtain development consent from this department 

before clearing native vegetation. There is provision 

for the development of Regional Vegetation 

Management Plans by government-appointed 

Regional Vegetation Management Committees with 

broad local representation. Where such a plan has 

been developed and approved, clearing may proceed 

in accordance with the plan without consent 

needing to be obtained for each instance. In districts 

such as Bega Valley Shire where no such plan has 

been developed, clearing consent provisions similar 

to the preceding legislation, State Environmental 

Planning Policy 46, still apply (that is, clearing 

applications are assessed on a property-by-property 

basis). 

There are numerous exemptions to the need to 

obtain approval for clearing under the NVC Act. 

Landowners may clear without approval up to two 

hectares per property per year and may cut up to 

seven trees per hectare per year for on-farm uses. 

Regrowth under 10 years old is also exempt, where 

the land was previously cleared. Forestry activities on 

public land and selective logging on private land are 

exempt, as is clearing authorised under the Noxious 

Weeds Act 1993, the Rural Fires Act 1996 and that 

required for the maintenance of powerline easements 

and road reserves. It can be seen from this that the 

NVC Act really only constrains large-scale clearing, 

and that it is inadequate to protect vegetation types 

that occur in small fragmented patches, most of 

which could be cleared under the exemptions. An 

opportunity will arise to address this problem with 

the drafting of a Regional Vegetation Management 

Plan for the local area. 

The NVC Act also provides for conservation of 

native vegetation on private property through the 

Native Vegetation Incentive Fund. In the Eden 

Management Area it is envisaged that this fund will 

provide one million dollars per year over two years. 

It is intended to provide incentives by way of 

funding for fencing of native vegetation to control 

stock access, with property owners signing Property 

Agreements with the Department of Land and 

Water Conservation undertaking to manage the 

relevant area of land for conservation outcomes. The 

intention of the Eden Region Forest Agreement is 

for such funding to be applied to conservation of 

‘ecosystems that are rare or non-existent on public 

lands’. That is, to the vegetation types listed in 

Table 5 (p. 52). 

4.2 Threatened Species 

Conservation Act 1995 

The Threatened Species Conservation Act 1995 (TSC 

Act) requires assessment of possible impacts of 

proposed developments where there is some 

indication that species of plants or animals listed as 

endangered (Schedule 1) or vulnerable (Schedule 2) 

in that Act may occur in the vicinity. If this 

assessment shows that the development could 

significantly affect a threatened species then the 

production of a Species Impact Statement is required 

to assist determination of whether the proposal 

should be allowed to proceed, be modified or 

refused. Where the determining authority for a 

development is not a departmental Minister, the 

concurrence of the Director-General of National 

Parks and Wildlife is required. In general the 

determining authority for most developments will be 

Council, which is then in the position of deciding 

whether an environmental impact assessment is 

required for particular development applications. 

73


Within Bega Valley Shire it is unlikely that 

threatened plant species will provide much 

impediment to development since (with three 

notable exceptions in the Lochiel area) most listed 

plant species occur predominantly or entirely within 

public lands. However, the presence of threatened 

fauna is more likely and may provide some 

protection to vegetation if it could be concluded that 

proposed clearing of vegetation would have a 

significant detrimental effect on local populations. 

The TSC Act is of little use in connection with 

depleted vegetation types found on agricultural 

lands, since none of the plant or animal species 

occurring in them in this area have been listed as 

threatened. However, it would be possible to 

nominate some of the relevant vegetation types as 

endangered communities under the TSC Act. If the 

nomination were accepted by the Scientific 

Committee, this would confer a greater degree of 

protection on remnant vegetation. Relevant 

vegetation types which occur only on private 

property are types 18, 20 and 21. There would be 

some precedent for nominating these vegetation 

types, since a similar vegetation type, Cumberland 

Plain Woodland (which occurs in western Sydney) 

has been listed. In the Australian Capital Territory, 

two communities found in agricultural areas have 

been listed as threatened under the Territory’s Nature 

Conservation Act 1980: Natural Temperate Grassland 

and Yellow Box – Blakely’s Red Gum Grassy 

Woodland. In Victoria, Forest Red Gum Grassy 

Woodland Community has been listed as threatened 

under the Flora and Fauna Guarantee Act 1988. 

There seems therefore to be a reasonable prospect 

that some or all of vegetation types 18, 20 and 21 

would be accepted for listing. Other communities, 

such as Dry Rainforest (1) and Floodplain Wetland 

(60), may also be likely to achieve listing since 

similar vegetation types, Herb-rich Plains Grassy 

Wetland (West Gippsland) Community and Dry 

Rainforest (Limestone) Community, are listed in 

Victoria. 

To accept an ecological community for listing as an 

endangered community, the Scientific Committee 

needs to be satisfied that: 

• the nominated community is sufficiently distinct 

from other species assemblages to be 

distinguished as a community; 

74 

• it is conventionally accepted as an entity; 

• there are threats to its continuing existence; and 

• it is likely to become extinct in nature in New 

South Wales unless the factors threatening its 

survival cease to operate. 

Of course in the case of a community rather than a 

species, the question of when it can be considered 

extinct is something of a grey area. From a perusal of 

information relating to listed communities in other 

areas, it appears that natural rarity or a high degree 

of clearing and fragmentation, with ongoing threat 

from agricultural activities, weed invasion or 

development pressure are the factors which generally 

prompt the acceptance of a nomination. However, 

the TSC Act exempts routine agricultural activities 

from requiring assessment of impact on listed 

communities, so listing would not hamper normal 

agricultural activities. The National Parks and 

Wildlife Service is required to prepare a Recovery 

Plan for listed communities, and listing of 

communities has in some cases been used to attract 

funding for regional conservation efforts. Any 

individual or body can make nominations. 

4.3 Fisheries Management Act 

1994 

The Fisheries Management Act 1994 provides some 

protection for estuarine vegetation in that a permit is 

required for any activity which might harm ‘marine 

vegetation’. At the moment this only includes sea 

grasses, mangroves and macro-algae (seaweed). Salt 

marsh and estuarine Melaleuca swamp or stands of 

sedges or reeds are not covered. A separate permit is 

required for dredging or reclamation. For minor 

proposals, New South Wales Fisheries would assess 

possible impacts, but for large-scale proposals an 

environmental impact assessment is required. Part 

7(a) of the Act covers threatened aquatic species, 

which are excluded from the Threatened Species 

Conservation Act 1995. Presently five fish species are 

listed, but no plants. The sea grass Posidonia australis 

may be proposed for listing in the future (R James, 

New South Wales Fisheries, pers. comm.). This 

species is recorded from Merimbula Lake and the 

Bermagui estuary (Keith & Bedward 1999).


4.4 State Environmental 

Planning Policy 14 

State Environmental Planning Policy 14 (SEPP 14) 

provides some protection for coastal wetlands. It 

requires landowners to obtain consent from councils 

before clearing, draining or filling wetlands, or 

constructing levees within them. The concurrence of 

the Director-General of the Department of Urban 

Affairs and Planning is also required. Environmental 

impact assessment is required to determine whether 

the proposed development would have a deleterious 

effect and whether there are feasible alternatives. 

However, SEPP 14 applies only to the wetland itself 

as mapped, not to its catchment, so the protection it 

provides is limited. It applies only to coastal 

wetlands. Some of the wetlands on the lower Bega 

River floodplain west of Kalaru are included, but 

none further inland. Even within the coastal strip 

some wetlands have been omitted from the listing of 

protected wetlands. SEPP 14 does not provide 

protection from grazing, which can be a major cause 

of degradation, nor address the problem of land uses 

in the wetland catchment which may cause erosion, 

sedimentation or weed invasion. 

4.5 State Environmental 

Planning Policy 44 

State Environmental Planning Policy 44 (SEPP 44) 

provides for koala habitat protection. It requires 

investigation of the presence of core koala habitat on 

any land which is the subject of a development 

application on which koalas could be expected to 

occur, from the presence of ‘potential koala habitat’. 

Core koala habitat is defined as land with a resident 

population of koalas, evidenced by attributes such as 

breeding females, and recent sightings of and 

historical records of a koala population. Potential 

koala habitat is defined as areas of native vegetation 

in which trees listed in Schedule 2 constitute at least 

15 per cent of the tree component. Of the trees so 

listed, only two occur in Bega Valley Shire, 

Eucalyptus tereticornis and E. viminalis. While 

E. tereticornis is assumed to have been an important 

feed tree at the time of European settlement of the 

district, koala populations have been eliminated 

from the areas where it occurs and it is not currently 

a significant species for local populations. 

E. viminalis is widespread but not particularly 

common in Bega Valley Shire and is unlikely to be 

significant for current known koala populations. 

Research by the South East Forests Conservation 

Council (South East Forests Koala Research Project 

1998), based on the distribution of koala faecal 

pellets under trees, has indicated that in Murrah and 

Bermagui State Forests the most used feed trees are 

Eucalyptus cypellocarpa and E. longifolia, with 

E. muelleriana, E. globoidea and E. bosistoana 

receiving a less significant level of use. No other area 

has been found in recent times in which there is a 

sufficient population of koalas to provide 

meaningful data about food preferences. On the 

information available it does not seem that SEPP 44 

provides any protection for either koalas or 

vegetation in this area, since investigation of 

‘potential koala habitat’ containing E. tereticornis or 

E. viminalis is extremely unlikely to result in the 

discovery of core koala habitat. Conversely, there 

may be areas in the Shire with the attributes of ‘core 

habitat’ (recent and historical sightings, presence of 

breeding females) which might go undetected 

because the absence of these two tree species means 

searches for koalas are not required. 

4.6 Department of Land and 

Water Conservation dam 

licensing requirements 

The construction of any dam for irrigation rather 

than for stock and domestic needs requires the 

issuing of a licence by the Department of Land and 

Water Conservation. Any dam of over seven 

megalitres capacity constructed on a ‘river’ for any 

purpose requires a licence (a river being defined as 

having a bed and banks, as opposed to a drainage 

line). For dams over 100 megalitres capacity, an 

environmental impact assessment is required, while 

for those over 50 megalitres, a statement of 

environmental effects is usually requested, although 

not mandatory. 

For dams which are not on rivers, the rules relating 

to impact assessment changed on 1 January 1999. 

Landowners are now only entitled to construct dams 

to harvest up to 10 per cent of the run-off from their 

property without requiring a licence, with some 

75


exceptions (for example, erosion control). How this 

is to be implemented in practice has yet to be 

determined. It may mean an improvement in 

environmental impact assessment of dam building in 

the district, since previously only dams on rivers 

required any assessment. As wetlands within 

agricultural areas have arguably been more severely 

depleted than any other vegetation type, this change 

is to be applauded. 

4.7 Bega Valley Local 

Environmental Plan 

The Bega Valley Local Environmental Plan 1987 

provides a framework for managing development 

within the Shire in such a way as to minimise 

damage to the environment by defining zonings and 

the developments which are permissible in them. 

The Local Environmental Plan sets out zones for 

environmental protection, but many of the more 

sensitive vegetation types and areas identified by the 

CRA, which occur principally on private lands, are 

not covered by any protective zoning. 

Techniques available to Council to improve 

protection of significant vegetation include: 

• extension of environmental protection zonings 

along riverine corridors; 

• definition of further areas of high scenic quality 

for scenic protection zoning; and, possibly 

• delineation on Local Environmental Plan maps 

of areas within the general agricultural zones 

where sensitive vegetation exists and where 

additional development assessment requirements 

might apply. 

A further technique for implementing the 

recommendations of this report could be the 

production of a Rural Lands Development Control 

Plan which: 

• specifies requirements for subdivision and other 

development; and 

• includes incentives and other positive measures 

for landowners to guide development to those 

parts of their properties where impacts on native 

vegetation and habitat are less significant. 

76 

This issue of course needs to be balanced against that 

of protection of prime agricultural land for 

agricultural uses. 

4.8 Extension programs and 

incentive schemes 

While the legislation and policy outlined above 

provides the ‘stick’ with which to prevent 

landholders from inflicting ecological damage which 

the community now deems unacceptable on the 

land in their care, there is much to recommend the 

use of ‘carrots’ to achieve the same aims. Tools 

generally include funding assistance and the 

provision of management advice. At present there 

are two schemes offering assistance to landholders 

with management of native vegetation. There is the 

Native Vegetation Incentive Fund discussed in 

section 4.1 above, which is administered by the 

Department of Land and Water Conservation. This 

scheme provides funds towards fencing costs and 

results in the signing of a Property Agreement 

between the landholder and the Department. 

However, this agreement does not result in a 

covenant on the property title. The National Parks 

and Wildlife Service administers a similar scheme 

which does involve placing a covenant on the title 

whereby Voluntary Conservation Agreements are 

signed by landholders and the Director-General of 

NPWS. Again, funding may be available for fencing 

and weed control. Both these schemes appear to be 

under-resourced in New South Wales, with no new 

staff being provided in local offices to oversee their 

implementation. The Voluntary Conservation 

Agreement scheme run by NPWS has a substantial 

backlog of unprocessed local applications (S Dovey, 

NPWS, pers. comm.). 

Another substantial source of funding for vegetation 

management projects in the district has been 

through grants administered by the Commonwealth 

Government under the Natural Heritage Trust, such 

as: 

• the One Billion Trees program; 

• the Save the Bush program; 

• the national Landcare program; and 

• the Rivercare program.


Grants have been made to local groups to fund such 

activities as fencing, erosion control and 

establishment of native vegetation. Most local efforts 

have been concentrated on riparian areas with the 

main aim of reducing erosion and sedimentation in 

streams. In the past, work on projects of this sort has 

often been hampered by the difficulty of obtaining 

appropriate technical advice. The situation has 

improved in recent years with more relevant, 

locally-based extension staff now available. However, 

these positions are themselves funded by the Natural 

Heritage Trust and have no long-term assurance of 

continuity. Also, staff are expected to service large 

numbers of community groups or large geographical 

areas, so the situation is less than ideal. 

77


5 Threatened flora in 


5.1 Species formally listed as 

rare or threatened 

There are several categories of rare or threatened 

species: 

• those listed as endangered or vulnerable at 

national level (Commonwealth Endangered 

Species Protection Act 1992); 

• those listed as endangered or vulnerable at State 

level (Threatened Species Conservation Act 1995); 

and 

• those listed as nationally rare by Briggs and Leigh 

(1995). 

This latter group, Rare or Threatened Australian 

Plants (ROTAP), includes species with restricted 

distributions which are not listed as threatened 

under national or State legislation because all or 

most of their distribution lies within conservation 

reserves. 

Species occurring in the Shire which are listed as rare 

or threatened in Commonwealth or State legislation 

or as ROTAPs are listed in Appendix 5 (p. 133). The 

list is taken from Table 3 in Keith, Miles and 

McKenzie (1999). However, species which occur 

within the Eden Management Area but outside the 

boundaries of Bega Valley Shire have been omitted, 

as have species whose only occurrence within the 

Shire is in Wadbilliga National Park. There are 16 

species in the latter category. Many of those 

remaining are known only from restricted locations 

within National Parks or other reserves (for example, 

Eucalyptus imlayensis from Mount Imlay and the 

orchid Prasophyllum affine from Nadgee). 

There is a group of species which are known only 

from rhyolite outcrops, mainly located west of 

Pambula in the South East Forests National Park 

(Yowaka Section), Nullica State Forest, Jingera Flora 

Reserve and adjacent private property. These include 

the listed species Genoplesium rhyoliticum, Leionema 

78 

ralstonii, Pseudanthus divaricatissimus, Westringia 

davidii, Zieria sp. G (buxijugum ms), Zieria sp. H 

(formosa ms) and Zieria sp. I (parrisiae ms). There 

are several other species which are largely confined to 

these rhyolite outcrops which have not been listed as 

rare or threatened, since outside the district they 

have a wider distribution. Another species, Acacia 

constablei, is known only from rhyolite outcrops in 

Narrabarba Hill Flora Reserve within Nadgee State 

Forest. Most of these species are endemic to Bega 

Valley Shire. That is, they are not known to occur 

anywhere else. Other Bega Valley Shire endemics 

which are not restricted to rhyolite are Acacia 

georgensis and Eucalyptus imlayensis. One other 

species, Acacia subporosa, is almost endemic to Bega 

Valley Shire, being known only from a restricted area 

in East Gippsland outside the Shire. However, 

within the Shire it is quite common along the coast, 

and is not listed as threatened. 

There are a small number of threatened species ‘hot 

spots’ which are not within the reserve system in the 

Shire, and which require alternative means of 

protection, such as within a Development Control 

Plan. These are: 

• Rhyolite outcrops on private property at Lochiel, 

where the only known populations of two 

shrubs, Zieria sp. G (buxijugum ms) and Zieria 

sp. H (parrisiae ms) occur. The populations are 

small and heavily browsed by wallabies, but do 

not appear to be threatened by livestock 

browsing, since they are in remote areas of the 

property which are fenced off from stock. Feral 

goat access from nearby forest is potentially a 

problem. 

• Two small rural residential blocks at Lochiel 

where a third endemic shrub, Zieria sp. H 

(formosa ms) occurs. This population is also 

small, and is growing on a rocky outcrop close to 

a dwelling where it is potentially vulnerable to 

damage from grazing, fire, weed infestation or 

accidental or deliberate destruction. Drought in


early 1998 appears to have killed a number of 

plants. All three Zieria species are listed as 

endangered (TSC Act) and their position 

currently appears to be very precarious. 

• Dr George Mountain trig station, where several 

listed species occur in close proximity, at a site 

which appears to receive quite a lot of public 

visitation. The species are Acacia georgensis 

(vulnerable on both Commonwealth and State 

threatened species schedules) and the ROTAP 

species Eucalyptus spectatrix, Leionema 

carruthersii, Haloragodendron bauerlenii and 

Pomaderris virgata. There are also a few 

individuals possibly belonging to an undescribed 

species of Pomaderris nearby (Neville Walsh, 

Melbourne Royal Botanic Gardens, pers. 

comm.). The other main population known of 

Acacia georgensis, at Kianinny Bay, is also in a site 

with a high rate of public visitation, making this 

species vulnerable to deliberate or accidental 

damage, although at Dr George it appears to be 

regenerating freely as a result of disturbance. The 

other species however, appear to have been badly 

affected by the drought in early 1998. This site is 

under assessment for inclusion in the adjacent 

Mimosa Rocks National Park, subsequent to the 

Eden Region Forest Agreement, but it appears 

unlikely that it will be included (P Cope, 

Narooma NPWS, pers. comm.). If this is the 

outcome, then management of the site is 

Council’s responsibility, as the site was 

apparently transferred from a trig reserve to 

Public Recreation Reserve (No. 85476) in 1965 

(D Hamer, Bega DLWC, pers. comm.). The site 

carries infrastructure associated with Council’s 

radio communication system, with overhead 

powerlines servicing it and a four-wheel drive 

access track. There has been rubbish dumping on 

the site. It would be preferable for the access 

track to be closed to discourage public use of the 

area. Fire regimes for the site also need to be 

addressed as local bush fire brigades have 

undertaken fuel reduction burns which have 

impinged on the area. African lovegrass is 

becoming established on the site. A management 

plan needs to be developed for the site in 

consultation with the National Parks and 

Wildlife Service. There is a draft recovery plan 

being developed for Acacia georgensis which is 

likely to provide management recommendations. 

The National Parks and Wildlife Service is required 

to produce a recovery plan for all threatened species. 

To date, recovery plans have nominally been 

produced for two species in Bega Valley Shire: Viola 

cleistogamoides and Pseudanthus ovalifolius. Both of 

these species are known from only a single record in 

the Shire, from Wonboyn (1954) and Ben Boyd 

National Park (1978) respectively. Despite survey 

efforts, populations have not been relocated. The 

recovery plans recommend further searches 

(K McDougall, NPWS, pers. comm.). Recovery 

plans are in progress for Eucalyptus imlayensis and 

Acacia georgensis and for several of the rhyolite 

species (the three Zieria species from Lochiel 

discussed above, Acacia constablei, Westringia davidii, 

Leionema ralstonii and Genoplesium rhyoliticum). 

5.2 Regionally uncommon 

species 

In addition to the species formally recognised as 

being rare or threatened, there are a considerable 

number of plant species found in Bega Valley Shire 

which are listed as being uncommon within the 

Eden Management Area, and a few which are 

uncommon throughout their distribution (Keith & 

Ashby 1992; Keith, Miles & McKenzie 1999). Some 

of these are known from only one record in the 

region, and may in fact be more threatened in the 

area than plants which are formally recognised as 

threatened. A high proportion of the listed 

threatened species occur entirely or largely within 

National Parks. Many of the uncommon species do 

also, but quite a high proportion of them are found 

only or mainly on private property, where it can be 

assumed their abundance has been reduced by 

clearing or agricultural activities. Some of these 

species are more common in other parts of their 

range, but some, because of a habitat preference for 

areas which have been most affected by agriculture, 

are uncommon or declining throughout their range. 

Species which are of regional conservation 

significance because they are regionally uncommon, 

or because their habitat has been depleted by 

clearing and agriculture, are listed in Appendix 6 

(p. 136). Some species are uncommon in the district 

because they are at the limit of their distribution 

here, and may only occur within a restricted location 

within the Shire. A considerable number of species, 

79


particularly rainforest species, are at their southern 

limit in the district, but species at their limit of 

distribution have not been listed in Appendix 6 

unless they are also regionally uncommon. Of the 

106 species which are regionally uncommon: 

• 45 occur primarily within the agricultural areas; 

• 29 are found in wetland or riparian habitats, 

often within agricultural areas; 

• 19 are more likely to occur in forests around the 

margins of the agricultural areas, but known 

occurrences are at least as common on private 

property as in public lands; and 

• 13 have a mainly coastal distribution where they 

are known to occur on private property as well as 

in reserves. 

80 

Species which are merely regionally uncommon 

rather than formally listed as threatened have no 

legislative protection. However, if one or more of the 

three most depleted Dry Grass Forest ecosystems of 

the Bega valley were to be listed as Endangered 

Ecological Communities under the Threatened 

Species Conservation Act 1995, the situation would 

improve for many of them, since 40 or so are mostly 

associated with vegetation types 18, 20 or 21 (Brogo 

Wet Vine Forest, Bega Dry Grass Forest and 

Candelo Dry Grass Forest).


6 Conclusions 

The CRA vegetation modelling has provided 

information about which vegetation types are most 

depleted and most threatened in the Shire and gives 

a general guide to where they might be expected to 

occur. However, the CRA vegetation map is not 

sufficiently accurate at fine scale to be used as a 

substitute for site inspection when particular sites are 

under consideration. 

The map also does not provide information about 

vegetation condition. It predicts which vegetation 

type would have occurred in which area pre-1750, 

then assumes that if the woody/non-woody layer 

indicates woody vegetation is still present on a 

particular site, that it is the same vegetation type 

which was there prior to European settlement. The 

exception to this is in the case where there is dense 

wattle scrub. In the map version which was assessed 

for this report, dense wattle scrub (usually Acacia 

mearnsii, black wattle regeneration) is distinguished 

from the eucalypt forest types (see Appendix 3, 

p. 122). On Version 4 of the map this distinction 

has been removed and vegetation which appears as 

wattle scrub on air photos is mapped as whatever 

eucalypt forest type the vegetation model predicts 

for the site. While wattle regeneration is still 

potentially of greater conservation significance than 

no native vegetation, it would generally be of less 

significance than eucalypt-dominated woody 

remnant vegetation. Council might therefore prefer 

to retain the old map version as well as obtaining 

Version 4, in order to have access to mapping which 

identifies wattle regeneration. 

It should also be remembered that the CRA extant 

vegetation map includes very little non-woody 

remnant vegetation within agricultural areas. 

Version 4 will include some areas of wetland 

vegetation, but possibly still omit numerous smaller 

remnants. Areas of low tree density (woodland) 

which may still have native grassy understorey of 

reasonable quality have been omitted. There may 

also be some areas of secondary grassland resulting 

from the clearing of trees from grassy open forests 

which are of conservation significance due to the 

presence of regionally uncommon grass and herb 

species, and which constitute remnants of vegetation 

types 20 or 21. About 40 such species, which are 

confined largely to remnants of vegetation types 

18–21, occur in the Shire. While natural grasslands 

are unlikely to have occurred in the Shire prior to 

European settlement, secondary grasslands or 

woodlands may now provide opportunities for 

conservation of these plant species. There has been 

no mapping of such vegetation in the Shire yet. 

The CRA vegetation work indicates that the most 

depleted vegetation types in the Shire are, with one 

exception, those associated with agricultural areas. 

Some are completely confined to such areas (18, 20, 

21), while others may extend outside them: 

• 19 (Bega Wet Shrub Forest) into surrounding 

public lands near drainage lines to a slight extent; 

• 39 (Northern Riparian Scrub) up river systems 

beyond the limit of clearing; 

• 60 (Floodplain Wetland) in drainage lines in 

uncleared coastal forests. 

However, the extent of occurrences outside private 

property is very limited and the reservation status of 

all these vegetation types very poor. Because they are 

fragmented and generally in small stands, 

conservation by acquisition into the reserve system is 

generally not a viable option. Conservation goals 

will need to be achieved by offering incentives to 

landowners to manage vegetation for conservation 

outcomes, and providing information to help them 

achieve this. Legislative constraints to clearing of 

remaining vegetation of depleted types may also 

have a place. Those constraints currently in existence 

(Threatened Species Conservation Act 1995, Native 

Vegetation Conservation Act 1997 and others 

discussed in Part B, section 4) do not provide 

adequate protection for the most depleted vegetation 

types. 

There are a number of poorly reserved vegetation 

types which are not associated with agricultural areas 

in the Shire. Generally these are types which are 

naturally rare, often through being confined to 

particular topographic situations such as drainage 

lines. Most have their non-reserved stands located 

largely within State Forests. An outcome of the Eden 

81


Region Forest Agreement has been the delineation of 

a network of informal reserves which exclude 

logging and provide for vegetation in designated 

areas to be managed for conservation under 

approved management plans (to be produced by 

April 2001). These informal reserves are located 

principally in the southern hinterland and protect 

Hinterland Heath (56) and Swamp Forest (58) 

vegetation types. Rainforests (types 6, 7 and 8) will 

continue to be managed within State Forests by 

prescription. 

Dune Dry Shrub Forest (36) is one under-reserved 

and naturally rare vegetation type which occurs 

mostly on private lands, but outside the agricultural 

areas. It is restricted to small patches close to the sea, 

or adjacent to coastal lakes (for example, Wallagoot 

Lake). Ground truthing of the mapping for this 

vegetation type showed that it is in fact even less 

extensive and less well reserved than the CRA model 

suggests. Main occurrences are between the southern 

outskirts of Bermagui and Cuttagee Lake, and 

between the Nullica River and Boydtown. The 

former occurrence is at least partially within vacant 

crown land. 

Threats to vegetation were discussed. Where 

vegetation has not already been cleared on private 

property, the principal threat is clearing and 

fragmentation as a result of residential development. 

Further degradation of remaining forest as a 

consequence of nearby clearing can be expected due 

to altered run-off patterns and nutrient enrichment, 

weed invasion and localised dieback caused by bell 

miners. In coastal areas most of the vegetation 

potentially subject to development pressures belongs 

to types which are regarded as adequately 

represented within the reserve system. However, 

development impacts on fauna populations may be 

significant, particularly where clearing would result 

in the attenuation or severing of forested links which 

could serve as corridors for fauna movement 

(‘wildlife corridors’). 

82 

The possible impacts of integrated logging on both 

public and privately owned forests were reviewed. 

These could include changes to species composition 

resulting from logging in some forest types, or from 

fire regimes, in addition to the obvious changes to 

age structure of the forests which may make them 

more vulnerable to severe wildfire damage and less 

able to support viable populations of 

forest-dependent fauna. The interaction between 

clearing or logging and bell miner dieback needs to 

be investigated in the region, as work in central and 

northern New South Wales suggests that bell miners 

are having an appreciable impact on forest health, 

particularly on disturbed forest margins. 

In the agricultural areas clearing is still potentially a 

threat to remaining remnants, and one which 

current legislation for the protection of native 

vegetation (Native Vegetation Conservation Act 1997) 

does very little to prevent, because of the range of 

exemptions which apply. Generally, however, the 

trend in Bega Valley Shire is towards regeneration of 

native vegetation in agricultural areas rather than 

further clearing. Regeneration is not uniformly 

distributed though, and some areas have very poor 

native vegetation cover. Continuing degradation is a 

threat to the persistence of existing vegetation. 

Causes include livestock access, dieback, lack of 

regeneration, invasion by agricultural weeds and by 

environmental weeds introduced through increased 

residential development in agricultural areas. Loss of 

genetic diversity through fragmentation of existing 

vegetation or through planting of local species 

derived from non-local seed could ultimately affect 

the viability of remnant vegetation.


7 Recommendations 

7.1 For Bega Valley Shire 

Council 

The highest priority for protection of remaining 

native vegetation should be allocated to remnant 

vegetation within agricultural areas, and to 

vegetation types which are depleted or naturally rare 

and under-reserved, which may occur outside 

agricultural areas but still largely on private property. 

Examples of the latter are Dune Dry Shrub Forest 

(36) on the coast and Bega Wet Shrub Forest (19) in 

the area between Wyndham and Burragate. Most 

common coastal vegetation types likely to occur in 

areas subject to development pressure are adequately 

reserved but may still be of conservation significance 

because they support threatened fauna species or 

contribute to potential wildlife corridors. 

Development proposals which involve clearing in 

areas which have been identified as potential wildlife 

corridors should be subject to specific consideration 

of the habitat values for fauna. These areas are: 

• vacant crown land between Pambula and 

Merimbula, and north of Merimbula; 

• the Bald Hills area west of Pambula and an area 

west of South Pambula; 

• an area of vacant crown land and private 

property straddling the Princes Highway 

between Wolumla and Merimbula; 

• the area of private land around the western side 

of Wallagoot Lake and between Bournda Nature 

Reserve and the northern section of Bournda 


• private property and vacant crown land between 

Kalaru and Tathra, through to the Bega River; 

• forested private property around Rileys Road 

between Cobargo and Bermagui which links the 

northern part of Murrah State Forest (now 

National Park) and Wallaga Lake National Park; 

• lower parts of the Towamba valley between 

Towamba and Kiah; 

• upper parts of the Towamba valley between New 

Buildings and Burragate; 

• the area along Burragate Road between 

Burragate and Wyndham; 

• the Yuranglo Voluntary Conservation Agreement 

area at Devil’s Hole south of Myrtle Mountain; 

• much of the Brogo area between Wadbilliga 

National Park in the west and Biamanga 

National Park in the east; 

• a patchy chain of remnants running from Brogo 

through Springvale to Black Range (Bournda 

Nature Reserve), spanning otherwise cleared 

agricultural areas; 

• a similar chain of remnants between the southern 

edge of South East Forests National Park, 

Bemboka and Tantawangalo Sections, running 

along the Bemboka River and Tantagwangalo 

Creek, though not confined to the riparian zone; 

and 

• the area between the former Murrabrine State 

Forest (now National Park) and Biamanga 

National Park, where privately owned forest 

almost bridges cleared country in the vicinity of 

Quinlans Road. 

Ultimately, Regional Vegetation Management Plans 

are intended to provide guidelines for conservation 

of remnant vegetation, identifying the location of 

those vegetation types which are depleted or 

threatened. Application would then need to be 

made, and approved by the Department of Land and 

Water Conservation, before clearing could proceed 

in these vegetation types. There is some scope here 

for reducing the exemptions in the case of particular 

vegetation types, to provide better protection for 

vegetation types which are reduced to small 

fragmented remnants. Although much of the 

groundwork for a Regional Vegetation Management 

Plan has already been done for this district (in the 

form of the CRA vegetation mapping and this 

report), it is still likely to be some time before such a 

plan is produced, since a Regional Vegetation 

Committee has yet to be appointed. The process 

could also be slowed down if the region, as defined 

for development of a Regional Vegetation 

Management Plan, includes areas outside the Eden 

83


Management Area, for which less detailed 

information will be available. The Southern Region 

CRA was completed in 2000 but little information 

has been made publicly available as yet. 

An alternative, or complementary, means of 

providing some protection for the most depleted 

vegetation types in agricultural areas is the listing of 

these as endangered ecological communities under 

the Threatened Species Conservation Act 1995. This 

has already been done: 

• in the Australian Capital Territory for remnant 

grassland and grassy woodland; 

• in Victoria for Forest Red Gum Grassy 

Woodland Community in East Gippsland; and 

• for Cumberland Plain woodland and several 

other vegetation types found in and around 

Sydney. 

Such listing would have no effect on routine 

agricultural activities, but would impose a 

requirement for environmental impact assessment 

on other developments. The vegetation types most 

likely to be successfully listed are 21 (Candelo Dry 

Grass Forest), 20 (Bega Dry Grass Forest) and 18 

(Brogo Wet Vine Forest), though others such as 1 

(Dry Rainforest) and 60 (Floodplain Wetland) could 

also be nominated. Existing mapping would need to 

be refined, since the CRA vegetation model provides 

only a rough indication of the location of these 

vegetation types, with many errors at the fine scale. 

For listing of such vegetation types as endangered 

communities to be applied in assessing development 

applications in the Shire, each site would need to be 

assessed individually to determine whether it in fact 

carried the relevant vegetation types, rather than 

relying on the CRA vegetation map. 

In the shorter term, a Development Control Plan 

would provide a vehicle for regulating development 

in environmentally sensitive areas within agricultural 

lands. The CRA extant vegetation map provides a 

rough guide for mapping environmentally sensitive 

areas, if one assumes that these are defined by the 

presence of vegetation types listed in point 1 below, 

or by the occurrence of potential wildlife corridors. 

While the fine detail about which vegetation types 

occur where may not be completely accurate, it is 

probably safe to assume that any remnant woody 

84 

vegetation within the core agricultural areas will be 

potentially of a significant type. The exception is 

wattle regeneration patches, which have been 

mapped separately from eucalypt-dominated 

remnants in the earlier version of the CRA map. 

Council should retain this earlier version because it 

provides this distinction. However, it would also be 

desirable to obtain the latest version, because of the 

improved mapping of some vegetation types (for 

example, 40, 60 and 1) in agricultural areas. 

To accurately map the location and type of all 

significant remnant vegetation in agricultural areas 

would be a major undertaking. It would probably be 

more efficient to assume that significant vegetation 

could occur anywhere within the agricultural areas 

and provide Council staff with training in 

recognising such vegetation, so that particular sites 

can be assessed as the need arises. 

Factors which need to be taken into account in 

assessing the conservation significance of remnant 

vegetation are: 

1. Vegetation type. The most significant types are: 

– Dry Rainforest (1); 

– Brogo Wet Vine Forest (18); 

– Bega Wet Shrub Forest (19); 

– Bega Dry Grass Forest (20); 

– Candelo Dry Grass Forest (21); 

– Northern Riparian Scrub (39); 

– Riverine Forest (40); 

– Floodplain Wetland (60) in agricultural 

areas; 

– Warm Temperate Rainforests (6 or 7); 

– Dune Dry Shrub Forest (36); 

– Floodplain Wetland (60); and 

– estuarine types 63–66 on the coast. 

2. Diversity of native species. Given that the 

grassy forest types carry a high proportion of 

their species diversity in the ground cover layer, 

this can really only be determined by someone 

with a thorough knowledge of local flora. 

Wetlands present a similar degree of difficulty in


distinguishing native from introduced flora. 

Generally it is safe to assume that the higher the 

native species diversity on a site, the less 

degraded is the remnant. However, some plant 

communities (for example, Phragmites reed beds) 

have a naturally low level of species diversity, but 

are still of conservation significance. 

3. Degree of weed invasion. Generally the lower 

the proportion of introduced species, the higher 

will be the conservation significance of the 

remnant. However, some level of introduced 

flora presence is inevitable in remnant 

vegetation, and a higher weed level could be less 

important than, say, the presence of significant 

species. 

4. Presence of regionally significant species. The 

presence of any of the species listed in 

Appendix 6 (p. 136) would increase the 

significance of any remnant stand. The greater 

the number of such species present, the greater 

the conservation significance. Plant species 

actually listed as rare or threatened in 

Commonwealth or State threatened species 

legislation are unlikely to occur on private 

property in Bega Valley Shire, although the 

possibility should not be discounted altogether. 

5. Structural diversity. Stands of greater structural 

diversity (for example, with several vegetation 

layers present, more than one age class of tree, 

litter and woody debris present) will generally be 

of more value to fauna and are likely to be in 

better health than stands with a simpler 

structure. However, stands with only one or two 

layers, such as Phragmites reed beds, secondary 

grassland or eucalypt woodland with a grassy 

understorey, may still be of high conservation 

significance. 

6. Age structure of the stand. The presence of 

more than one age class of tree is desirable. Old 

trees and dead stags provide habitat for 

hollow-dependent fauna. Mature trees provide 

for continuity of the supply of hollows as well as 

foraging and nesting habitat. Regenerating trees 

provide for the long-term persistence of the 

stand. 

7. Isolation or contiguity of the stand. In an area 

where remnant vegetation is abundant, one 

particular stand may be regarded as expendable, 

depending on how it scores on the other factors 

discussed here. Conversely, in an area where 

remnant vegetation is sparse or degraded, each 

stand is important, so that stands of lower 

quality might be accorded a higher conservation 

significance than they would elsewhere. A stand 

which is strategically located so as to be a 

potential wildlife corridor will be of greater value 

than an isolated stand, all other things being 

equal. Stands which are connected to other 

vegetation are also more likely to remain viable 

in the long term. 

8. Size of stand. In general, the larger the better. 

Very small stands are unlikely to sustain the 

ecosystem processes necessary to keep them 

viable. However, given a reasonable level of 

species diversity, small stands can be gradually 

enlarged by encouraging natural regeneration or 

by buffer plantings of appropriate species. 

Assessment of the conservation significance of 

remnant vegetation which occurs on publicly owned 

lands which potentially come under Council’s 

management would be valuable. This could include 

roadsides (for which some preliminary assessment 

has already occurred), cemeteries, showgrounds and 

sports grounds in the smaller towns and travelling 

stock reserves, which are managed by the Rural 

Lands Protection Board. Some valuable remnants of 

vegetation types confined to the agricultural lands 

could be found in these sites. A known example is 

the Tantawangalo speedway reserve, which carries 

riparian vegetation and a small remnant of Candelo 

Dry Grass Forest (21) in good condition. It would 

be valuable to provide training in remnant 

vegetation recognition and management to ground 

staff and community volunteers involved in 

managing such areas. The Recreation Reserve on the 

Dr George Mountain trig site particularly requires 

the development and implementation of a 

management plan by Council, in consultation with 

NPWS. This is due to the large number of rare or 

threatened species in the vicinity and the site’s 

vulnerability due to its ease of access. 

85


The Tree Preservation Order in force in the Shire 

should be reviewed to determine whether it can be 

used to protect remnant vegetation, since the Native 

Vegetation Conservation Act 1997 does not presently 

provide real protection for small fragmented 

remnants. Extension of the Tree Preservation Order 

to cover rural residential lots within agricultural 

zonings could be valuable. Properties which are 

deemed to be of a viable size for farming could be 

excluded. 

Council may be able to ameliorate the impacts of 

medium density residential development in coastal 

areas by attention to design features of subdivisions. 

Control of stormwater run-off to prevent it entering 

remaining natural vegetation is important in 

minimising weed encroachment in these areas, as is 

control of domestic waste water and septic effluent. 

It would be preferable to sewer such developments 

and to run both sewage and stormwater systems 

through artificial wetlands prior to release of treated 

water into the environment. The siting of 

subdivisions within a ring road, rather than backing 

housing lots directly onto forest, offers advantages in 

fire control as well as making it more difficult for 

residents to dump garden refuse in the bush (a 

common source of weed invasion in residential 

areas). 

Council should continue to support remnant 

vegetation conservation through support of projects 

such as the Remnant Vegetation Recovery Program. 

Another worthwhile project would be the 

production of a brochure discussing remnant 

vegetation, Landcare, erosion, weeds and similar 

issues, for distribution to all new landholders in the 

Shire. It would be desirable to resurrect this project, 

which was proposed by the Environment Network 

Centre but did not receive funding. Support for 

Greencorps projects tackling environmental 

problems in the district is also desirable. 

7.2 For the Catchment 

Management Committee 

The Catchment Management Committee is able to 

influence impacts on vegetation in the Shire largely 

through support for funding applications. The 

committee should continue to support and 

86 

encourage development of projects relating to 

conservation of remnant vegetation in agricultural 

areas. Within these areas riparian vegetation is often 

particularly important since it serves to protect 

streams from erosion as well as providing for 

conservation of flora and fauna. Riparian vegetation 

has been the main area addressed by local Landcare 

projects to date. 

The recommendation by Fryirs and Brierley (1998b) 

that greatest effort should be put into conserving 

existing riparian vegetation rather than trying to 

restore the most degraded areas is endorsed. These 

areas are often of high conservation significance as 

well as being strategically important for preventing 

erosion and the release of large volumes of sediment. 

Projects which are likely to contribute to the creation 

of substantial linkages between patches of remnant 

vegetation should receive support. 

In general it is preferable to encourage natural 

regeneration rather than to plant vegetation. This is 

because of the possibility of creating environmental 

weeds through planting inappropriate species, or of 

swamping local gene pools with the introduction of 

plants from non-local seed. Grant applications need 

to be scrutinised for appropriateness of planting lists, 

and groups should be encouraged to use natural 

regeneration where this is an option. On the less 

degraded sites which Fryirs and Brierley recommend 

addressing there is likely to be no need for planting. 

Landcare and Rivercare groups should be 

discouraged from planting trees merely in order to 

be seen to be doing something, if there is no 

biological necessity for it. 

When providing advice to landholders on plantation 

establishment, the Farm Forestry Project needs to 

take into account the possible impacts of use of 

exotic or non-local species (in terms of spread of 

environmental weeds) and of use of local species of 

non-local provenance. More consideration needs to 

be given to use of suitable local species and means by 

which their performance could be improved, rather 

than encouraging use of exotic and non-local species. 

The siting of plantations close to high 

conservation-value remnant vegetation would 

require careful consideration of potential impacts.


Remnant vegetation is frequently weed infested and 

control of weeds probably needs to be given a higher 

profile in Landcare activities. In many instances 

natural regeneration will occur with suitable site 

management, such as removal of weeds. However, 

different vegetation types will require different 

management strategies. Groups and landholders will 

need a solid information base if they are going to 

embark on such activities. Provision of a bush 

regeneration course with a local orientation through 

the local technical and further education (TAFE) 

college would be useful, as would subsidised courses 

in herbicide use for Landcare members. The 

appointment of an extension officer with secure 

tenure to provide advice on management of remnant 

vegetation is desirable. 

An environmental weed of particular concern in the 

Shire is the willow. The Catchment Management 

Committee has helped raise the profile of this 

problem, and mapping of willows and seedling 

infestation has occurred for most of the larger 

streams in the agricultural areas. However, action has 

so far been limited to the efforts of Landcare and 

Rivercare groups in relatively small areas. There 

needs to be development and implementation of a 

detailed willow control plan, which does not rely on 

the voluntary labour of a few committed individuals. 

This may necessitate the employment of a crew to 

control willows in accordance with the plan. A high 

priority should be given to willow removal in areas 

where they are still uncommon, and where the bulk 

of the riparian vegetation is native, especially in 

upstream sections of the river systems. 

A survey of the extent and impact of bell miner 

dieback across all land tenures in the region is 

urgently required. The Eden Region Forest 

Agreement suggests that assessment of bell miner 

effects should be part of ecologically sustainable 

forest management. 

7.3 For the Regional Native 

Vegetation Committee 

An outcome of the Eden Region Forest Agreement 

has been the requirement that a Native Vegetation 

Committee be appointed in the region (as required 

under the Native Vegetation Conservation Act 1997) 

by 30 June 1999. One of the main tasks of such a 

committee will presumably be to study the CRA 

vegetation information base and documents such as 

this report in order to make firm recommendations 

about the management of native vegetation within 

the region, particularly on private lands. 

Two points stand out as requiring clarification as a 

result of this report. One is how to determine exactly 

which remnant vegetation is of high conservation 

significance in the Shire, given that the CRA 

vegetation map is not highly accurate at fine scale, 

and that it does not address vegetation condition. 

The other is the issue of the maintenance of wildlife 

corridors across private lands between areas of native 

forest which are largely in public lands. The 

approximate location of such areas has been 

suggested in this report, but detailed mapping will 

be needed if the recommendation to more closely 

scrutinise development applications in these areas is 

to be workable. 

87


List of plant common names 

Common and scientific names of exotic species referred to in the text can be found in Appendix 4 (p. 125). For 

native species, only those with a common name that is genuinely in common usage are listed here. 

Acacia cognata bower wattle 

Acacia dealbata silver wattle 

Acacia falciformis broad-leaved hickory 

Acacia floribunda white sallow or sallee wattle 

Acacia implexa hickory, or lightwood 

Acacia longifolia Sydney golden wattle 

Acacia mearnsii black wattle 

Acacia melanoxylon blackwood 

Acacia silvestris Bodalla silver wattle 

Acacia sophorae coast wattle 

Acacia subporosa bower wattle 

Acacia terminalis sunshine wattle 

Acmena smithii lilly pilly 

Allocasuarina littoralis black she-oak 

Allocasuarina verticillata (formerly Casuarina stricta) drooping she-oak 

Angophora floribunda rough-barked apple, apple gum, apple box 

Atherosperma moschatum southern sassafras 

Backhousia myrtifolia grey myrtle 

Banksia integrifolia coast banksia 

Banksia serrata saw banksia 

Bedfordia arborescens blanketleaf, blanket bush 

Brachychiton populneus kurrajong, yam tree 

Bursaria lasiophylla, B. spinosa blackthorn 

Carex longebrachiata Bergalia tussock 

Cassinia aculeata, C. longifolia, C. trinerva dogwood 

Casuarina cunninghamiana river oak, river she-oak 

Casuarina glauca swamp oak 

Corymbia gummifera (formerly Eucalyptus gummifera) red bloodwood 

Corymbia maculata (formerly Eucalyptus maculata) spotted gum 

Dendrobium speciossum rock orchid, rock lily 

Dicksonia antarctica soft tree fern 

Doryphora sassafras sassafras 

Elaeocarpus holopetalus black olive-berry 

Elaeocarpus reticulatus blue olive-berry, blueberry ash 

Eucryphia moorei pinkwood, plumwood 

Eucalyptus agglomerata blue-leaved stringybark 

88


Eucalyptus angophoroides apple-topped box 

Eucalyptus baueriana blue box (locally also called black box) 

Eucalyptus baxteri brown stringybark 

Eucalyptus bosistoana coast grey box, Gippsland grey box 

Eucalyptus botryoides bangalay, southern mahogany 

Eucalyptus consideniana yertchuk 

Eucalyptus cypellocarpa monkey gum, mountain grey gum 

Eucalyptus dalrympleana mountain gum 

Eucalyptus elata river peppermint (also locally ribbony gum or 

white peppermint) 

Eucalyptus fastigata brownbarrel 

Eucalyptus fraxinoides white ash 

Eucalyptus longifolia woollybutt 

Eucalyptus maidenii Maiden’s gum, blue gum 

Eucalyptus melliodora yellow box 

Eucalyptus obliqua messmate 

Eucalyptus ovata swamp gum 

Eucalyptus paniculata grey ironbark 

Eucalyptus pauciflora snow gum, white sallee 

Eucalyptus pilularis blackbutt 

Eucalyptus polyanthemos ssp. vestita red box 

Eucalyptus radiata narrow-leaved peppermint (also locally black 

peppermint) 

Eucalyptus sieberi silvertop ash 

Eucalyptus smithii gully gum 

Eucalyptus stenostoma Jilliga ash 

Eucalyptus tereticornis forest red gum 

Eucalyptus tricarpa (formerly E. sideroxylon ssp. tricarpa) red ironbark 

Eucalyptus viminalis manna gum, ribbon gum 

Ficus rubiginosa rusty fig, Port Jackson fig 

Hedycarya angustifolia austral mulberry 

Hymenanthera dentata tree violet (also locally whitethorn) 

Imperata cylindrica blady grass 

Indigofera australis austral indigo 

Jacksonia scoparia dogwood 

Kunzea ambigua tickbush 

Kunzea ericoides (formerly Leptospermum phylicoides) burgan (also locally tea tree) 

Lomandra longifolia mat rush 

Melaleuca armillaris bracelet honey myrtle 

Melaleuca ericifolia swamp paperbark (also locally tea tree) 

Microlaena stipoides weeping grass 

Notelaea venosa mock olive 

Persoonia linearis narrow-leaved geebung 

89


Phragmites australis common reed 

Pittosporum undulatum sweet pittosporum 

Polyscias murrayi pencil cedar 

Pomaderris aspera hazel pomaderris 

Pteridium esculentum bracken 

Solanum aviculare kangaroo apple 

Tetrarrhena juncea wiregrass 

Themeda australis kangaroo grass 

Urtica incisa stinging nettle 

90


List of acronyms 

API air photo interpretation 

BVSC Bega Valley Shire Council 

CMC Catchment Management Committee 

CRA Comprehensive Regional Assessment 

DLWC Department of Land and Water Conservation 

EIS Environmental impact statement 

EFA Eden Region Forest Agreement 

EMA Eden (Native Forests) Management Area (of SFNSW) 

FCNSW Forestry Commission of New South Wales (now SFNSW) 

FR Flora Reserve 

FSC CMC Far South Coast Catchment Management Committee 

GIS geographic information system (i.e. maps on computer) 

IFA Interim Forest Agreement 

NP National Park 

NPWS NSW National Parks and Wildlife Service 

NR Nature Reserve 

NVC Act Native Vegetation Conservation Act 1997 

RFA Regional Forest Agreement 

ROTAP Rare or threatened Australian plant (a listing by Briggs & Leigh 1995) 

SEFNP South East Forests National Park 

SEPP State Environmental Planning Policy 

SFNSW State Forests of New South Wales (formerly FCNSW) 

sp. species (singular) 

spp. species (plural) 

ssp. subspecies 

TSC Act Threatened Species Conservation Act 1995 

91


Glossary 

alluvium material deposited by flowing water. Used loosely to include soils developed from such 

materials along existing and former watercourses. 

Devonian 

metasediments 

92 

generally coarse-grained sedimentary rock (conglomerate, sandstone and siltstones) 

deposited in a freshwater environment during the Devonian era (395–345 million years 

ago) and metamorphosed to varying degrees by contact with molten magma. A 

common rock type particularly in the southern half of the Shire, giving rise to sandy 

soils of low fertility. 

endemic of plants or animals, occurring only in a particular place. 

epiphytic of plants, growing on other plants but not drawing nourishment from them (for 

example, some orchids on trees, algae on seagrasses). 

forb a non-woody plant (that is, herb) which is not a grass or graminoid. 

forest, closed forest where canopy cover of trees shades over 70 per cent of the ground (for example, 

rainforest). 

forest, open forest where tree crowns shade 30–70 per cent of the ground. 

graminoid a non-woody plant similar in appearance to grasses, but not a grass (for example, sedges, 

rushes). 

granitoid a group of rock types formed from slow cooling of molten magma below the earth’s 

surface and having a relatively large crystal size. Includes granite, granodiorite, tonalite 

and adamellite, all of which occur locally, and are colloquially referred to as granite. 

heath a vegetation community dominated by sclerophyllous shrubs usually found on acid soils 

of low fertility such as those derived from sandstones. Wet heath, found along drainage 

lines or in seepage areas, also includes a high proportion of sedges. 

Holocene sands sand deposited in the Holocene (current) geological epoch commencing 10 000 years 

ago, and which has consequently not begun the transformation into rock. Generally 

coastal sand dunes or vegetated old dune systems (for example, opposite Merimbula 

airport). 

mesophyll of plants, soft-leaved, growing in situations with an adequate water supply (see 

sclerophyll). 

Ordovician 

metasediments 

fine-grained sedimentary rock (shales and siltstones) laid down under the sea in the 

Ordovician era (525–440 million years ago) and slightly metamorphosed by heat or 

pressure. Common in coastal areas, particularly north of Bega, and occasionally as 

screens and isolated outcrops within the granitoid rocks of the Bega valley. Derived soils 

are higher in clay, more fertile and have better water-holding capacity than those from 

Devonian metasediments. 

plutonic of rock, formed by slow cooling of large masses of molten magma beneath the earth’s 

surface.


rain shadow areas which receive lesser rainfall because they are in the lee of a range which intercepts 

rainfall are said to be in a rain shadow. The coastal range casts a rain shadow over 

adjacent inland valleys. 

relict of plants, an individual or population persisting in a habitat which has been changed 

from that which was typical for it. 

remnant 

vegetation 

a fragment remaining of an originally more widespread vegetation type. In this area 

such vegetation usually consists of regrowth following clearing, although it may include 

some individual relict trees which were present at the time of European settlement. 

rhyolite a fine-grained acid volcanic (formed by rapid cooling on the earth’s surface, 

c.f. plutonic) rock, which resists weathering and produces only shallow and infertile 

soils. Most common west of Pambula, but also occurring at Kianinny, Bunga Head and 

Narrabarba. 

riparian pertaining to the bed and banks of a river or creek. Generally not used for smaller 

drainage lines without defined banks. 

secondary 

grassland 

vegetation in which the predominant species are native grasses and forbs, formed by the 

clearing of trees from what was originally woodland or open forest with a 

predominantly grassy understorey. 

sclerophyll of plants, having hard and often small leaves, generally as an adaptation to reduce water 

loss, or prevent browsing by animals in response to low soil fertility. Typical of plants 

found on drier, less fertile soils. 

stratification the division of an area into different zones based on factors such as geology, rainfall and 

temperature gradients. Done in order to ensure that sampling of vegetation or fauna 

takes into account the full range of variation in such factors which might affect their 

distribution and can be expected to occur in the area. 

Tertiary 

sediments 

sediments laid down by freshwater along the coastline during the Tertiary period 

(65–2.6 million years ago) and giving rise to very infertile soils. Scattered along the 

coastline, but most common south of Eden. 

woodland vegetation type in which tree crowns shade only 10–30 per cent of the ground. On soils 

of moderate to high fertility, grasses are likely to dominate the understorey, while heathy 

shrubs are more likely to be dominant on very infertile soils. 

93


Bibliography and references 

Anon (1998) API (Aerial Photographic 

Interpretation) Mapping Report for Eden Region. 

Draft Report for NSW CRA/RFA Steering 

Committee. 

Anon (1999) Forest Agreement for Eden Region. 

New South Wales Government, 5 March 1999. 

Austin M.P. & Margules C.R. (1986) Assessing 

representativeness. In M.B. Usher (ed). Wildlife 

conservation evaluation, pp 45–67.(Chapman and 

Hall: London). 

AWT EnSight (1997) Baseline Ecological Study of 

the Bega River System. Report No. 97/251 prepared 

for Far South Coast Catchment Management 

Committee. 

Backhouse G.N. (1987) Management of Remnant 

Habitat for Conservation of the Helmeted 

Honeyeater Lichenostomus melanops cassidix. In D.A. 

Saunders, G.W. Arnold, A.A. Burbidge & A.J.M. 

Hopkins (eds) Nature Conservation: The role of 

remnants of native vegetation. (Surrey Beatty and 

Sons Pty. Ltd.: Chipping Norton). 

Banks J.C.G. (1990) The fire history for two forest 

types in Glenbog State Forest, NSW. Report to the 

Joint Scientific Committee on the South East 

Forests, Canberra. 

Bedward M., Pressey R.L. & Keith D.A. (1992a) A 

new approach for selecting fully representative 

reserve networks: Addressing efficiency, reserve 

design and land suitability with an iterative analysis. 

Biological Conservation 62: 115–125. 

Bedward M., Keith D.A. & Pressey R.L. (1992b) 

Homogeneity analysis: Assessing the utility of 

classifications and maps of natural resources. 

Australian Journal of Ecology 17: 133–140. 

Belbin L. (1994) PATN. Pattern analysis package 

(CSIRO: Canberra). 

Benson J.S. & Redpath P.A. (1997) The nature of 

pre-European native vegetation in south-eastern 

Australia: A critique of Ryan, D.G., Ryan, J.R. and 

Starr, B.J. (1995) The Australian landscape – 

observations of explorers and early settlers. 

Cunninghamia 5: 285–328. 

94 

Bertram R. & Allen C. (1998) South East Forests 

Koala Research Project. Unpublished report to 

South East Forests Regional Adjustment 

Committee. 

Bridges R.G. (1983) Integrated logging and 

regeneration in the silvertop–stringybark forests of 

the Eden region. Research Paper No. 2. Forestry 

Commission of NSW, Sydney. 

Brierley G.J., Fryirs K. & Cohen T. (1996) 

Geomorphology and river ecology in south-eastern 

Australia: An approach to catchment 

characterisation. Graduate School of the 

Environment, Macquarie University. Working 

Papers for LWRRDC Project MQU1. 

Brierley G. & Murn C.P. (1995) Geomorphic 

characterisation of Cobargo catchment. School of 

Earth Sciences, Macquarie University. Report to 

Cobargo Landcare Group. 

Briggs J.D. & Leigh J.H. (1988) Rare or Threatened 

Australian Plants, first edition, CSIRO Australia. 

Briggs J.D. & Leigh J.H. (1995) Rare or Threatened 

Australian Plants, second edition, CSIRO Australia. 

Brooks A. (1994) Vegetation and channel 

morphodynamics along the lower Bega River. 

Honours thesis, School of Earth Sciences, Macquarie 

University. 

Clark B. (undated) An analysis of water quality data 

of the far south coast. Unpublished report prepared 

for Water Quality Sub-committee of the Far South 

Coast Catchment Management Committee. 

Clarke M.F., Grey M.J., Britton D.R. & Loyn R.H. 

(1995) The noisy miner (Manorina melanocephala) 

and rural dieback in remnant eucalypt woodlands. 

RAOU Report No. 98. Royal Australasian 

Ornithologists Union, Melbourne. 

Clements A. (1983) Suburban development and 

resultant changes in the vegetation of the bushland 

of the northern Sydney region. Australian Journal of 

Ecology 8: 307–318.


Commonwealth of Australia (1992) National forest 

policy statement. A new focus for Australia’s forests. 

Second edition. Australian Government Publishing 

Service, Canberra. 

Crawford I. (1992) Vegetation mapping of 

Wadbilliga National Park and the proposed 

Bemboka National Park. Unpublished report 

prepared for NSW NPWS Queanbeyan Regional 

Office. 

de Chazal J. (1992) A preliminary comparative study 

of the effects of clear-felling and wildfire upon the 

floristic and structural composition of forest 

regeneration in the E. sieberi–stringybark forests of 

the Eden region, NSW. Honours thesis, School of 

Resource and Environmental Management, 

Australian National University. 

Department Of Land and Water Conservation 

(1997) Definitions and Exemptions, State 

Environmental Planning Policy No. 46 – Protection 

and Management of Native Vegetation. 

Dickinson K.J.M. & Kirkpatrick J.B. (1987) 

Short-term effects of clear-felling and slash-burning 

on the richness, diversity and relative abundance of 

higher plant species in two types of eucalypt forest 

on dolerite in Tasmania. Australian Journal of Botany 

35: 601–616. 

Felton-Taylor J. (1998) Pambula Estuarine Wetland 

Study. Unpublished report prepared for Department 

of Land and Water Conservation and the Pambula 

Area Planning and Progress Association. 

Floyd A.G. (1982) Rainforests of the south coast of 

New South Wales south of Batemans Bay. 

Unpublished report to NSW NPWS. 

FCNSW (1989) Forest types of New South Wales. 

Research Note 17. Forestry Commission of New 

South Wales, Sydney. 

FCNSW (1988) Forestry operations in Eden 

management area. Environmental Impact Statement. 

Forestry Commission of New South Wales, Sydney. 

FCNSW (1983) Notes on the silviculture of major 

NSW forest types. 2. Moist tableland Hardwood 

types. Unpublished report. 

Fryirs K.A. (1995) The character and evolution of 

valley fills in upper Wolumla Creek catchment, 

South Coast NSW. Honours thesis, School of Earth 

Sciences, Macquarie University. 

Fryirs K. & Brierly G. (1998a) River styles in 

Bega–Brogo catchment: Recovery potential and 

target conditions for river rehabilitation. School of 

Earth Sciences, Macquarie University. Report 

completed for Bega Valley Shire, Far South Coast 

Catchment Management Committee and 

Department of Land and Water Conservation. 

Fryirs K. & Brierley G. (1998b) The use of river 

styles and their associated sediment storage in the 

development of a catchment-based river 

rehabilitation strategy for Bega–Brogo catchment, 

South Coast, NSW. School of Earth Sciences, 

Macquarie University. Report completed for Bega 

Valley Shire, Far South Coast Catchment 

Management Committee and Department of Land 

and Water Conservation. 

Gibbons P. (1994) Sustaining key old-growth 

characteristics in native forests used for wood 

production: Retention of trees with hollows. In 

Norton T.W. & Dovers S.R. (eds) Ecology and 

Sustainability of Southern Temperate Ecosystems. 

CSIRO Australia. 

Gibbons P & Lindenmayer D.B. (1997) The 

performance of prescriptions employed for the 

conservation of hollow-dependent fauna: 

Implications for the Comprehensive Regional 

Assessment process. Working Paper 1997/2. 

Graham-Higgs N & Assoc. (1995) Report on the 

results of environmental controls for the 

development of the seaphone facility atop 

Mount Imlay. Prepared for Telstra, February 1995. 

Green D. (1999) A survey of wetlands in the Bega 

valley. Department of Land and Water 

Conservation, Sydney South Coast Region. 

Harden G. (1992) Flora of New South Wales. 

(University of New South Wales Press: Kensington). 

Harris-Daishowa (1986) Draft environmental impact 

statement for the Eden (NSW) export woodchip 

operation for the period 1989–2009. 

Harris-Daishowa, Canberra. 

Hussey BMJ & Wallace KJ (1993) Managing your 

bushland. Department of Conservation and Land 

Management, WA. 

95


Hutchinson MF (1989) A new objective method for 

spatial interpolation of meterological variables from 

irregular networks applied to the estimation of 

monthly mean solar radiation, temperature, 

precipitation and windrun. Technical Memorandum 

89/5 (CSIRO Division of Water Resources: 

Canberra). 

JANIS (1996) Proposed nationally agreed criteria for 

the establishment of a comprehensive, adequate and 

representative reserve system for forests in Australia. 

Report of the Joint ANZECC/MCFFA National 

Forest Policy Implementation Subcommittee 

(JANIS). Commonwealth of Australia, Canberra. 

Keith D.A. (1994) Conservation and management 

of natural vegetation in the Bega valley and adjacent 

areas. NSW National Parks and Wildlife Service. 

Unpublished report. 

Keith D.A. (1996) Involving ecologists and local 

communities in survey, planning and action for 

conservation in a rural landscape: An example from 

the Bega valley, New South Wales. In D.A. Saunders, 

J. L. Craig & E. Mattiske (eds) Nature 

Conservation 4: The role of networks (Surrey Beatty 

and Sons Pty Ltd: Chipping Norton). 

Keith D.A. (1996) Fire-driven extinction of plant 

populations: A synthesis of theory and review of 

evidence from Australian vegetation. Proceedings of 

the Linnean Society of New South Wales 116: 37–78. 

Keith D. A. & Ashby E. (1992) Vascular plants of 

conservation significance in the South East Forests of 

New South Wales. Occasional Paper No. 11. National 

Parks and Wildlife Service of New South Wales, 

Hurstville. 

Keith D. A., Bedward M. & Smith J. (1995) 

Vegetation of the South East Forests of New South 

Wales. NSW National Parks and Wildlife Service. 

Unpublished report. 

Keith D.A. & Bedward M. (1998) Forest Ecosystem 

Classification and Mapping for the Eden 

Comprehensive Regional Assessment. Unpublished 

report to NSW CRA/RFA Steering Committee. 

Keith D.A. & Bedward M. (1999) Native Vegetation 

of the South East Forests Region, Eden, New South 

Wales. Cunninghamia 6(1): 1–218. 

96 

Keith D.A., Miles J & Mackenzie B. (1999) The 

vascular flora of the South East Forests region, Eden, 

New South Wales. Cunninghamia 6(1): 219–281. 

Keith D.A. & Sanders J.M. (1990) Vegetation of the 

Eden Region, South-eastern Australia: Species 

composition, diversity and structure. Journal of 

Vegetation Science 1: 203–232. 

Landsberg J. & Wylie F.R. (1983) Water stress, leaf 

nutrients and defoliation: A model of dieback of 

rural eucalypts. Aust. Journal of Ecology 8: 27–41. 

Low T. (1994) Invasion of the savage honeyeaters. 

Australian Natural History 24(10): 27–33. 

Loyn R.H. (1987) Effects of patch size and habitat 

on bird abundances, species numbers and tree health 

in fragmented Victorian forest. In D.A. Saunders, 

G.W. Arnold, A.A. Burbidge & A.J.M. Hopkins 

(eds) Nature Conservation: The role of remnants of 

native vegetation. (Surrey Beatty and Sons Pty Ltd: 

Chipping Norton), pp 65–75. 

Loyn R.H., Fagg P.C., Piggin J.E., Morton A.G. & 

Tolhurst K.G. (1983) Changes in composition of 

understorey vegetation after harvesting eucalypts for 

sawlogs and pulpwood in East Gippsland. Aust. 

Journal of Ecology 8: 43–53. 

Lunney D. & Leary T. (1988) The impact on native 

mammals of land-use changes and exotic species in 

the Bega district, New South Wales, since 

settlement. Australian Journal of Ecology 13: 67–92. 

Lunt I.D. (1995) European management of remnant 

grassy forests and woodlands in south-eastern 

Australia – past, present and future? Victorian 

Naturalist 112: 239–249. 

Margules C.R. & Stein J.L. (1989) Patterns in the 

distribution of species and the selection of nature 

reserves: An example from the Eucalyptus forests in 

south-eastern New South Wales, Australia. Biological 

Conservation 50: 219–238. 

Marsh N.R. & Adams M.A. (1995) Decline of 

Eucalyptus tereticornis near Bairnsdale, Victoria: 

Insect herbivory and nitrogen fractions in sap and 

foliage. Aust. Journal of Botany 43: 39–50. 

Mueck S.G. & Peacock R.J. (1992) Impacts of 

intensive timber harvesting on the forests of East 

Gippsland, Victoria. Department of Conservation 

and Natural Resources VSP Technical Report No. 15.


Ough K. & Ross J. (1992) Floristics, fire and 

clear-felling in wet forests of the central highlands, 

Victoria. VSP Technical Report No. 11. Department 

of Conservation and Environment, Melbourne. 

Patterson, Britton & Partners (1996) Wallaga Lake 

Estuary Processes Study. A report commissioned to 

Department of Public Works South Coast on behalf 

of Bega Valley Shire Council, Eurobodalla Shire 

Council and Wallaga Lake Estuary Management 

Committee. 

Recher H.F., Shields J., Kavanagh R. & Webb G. 

(1987) Retaining remnant mature forest for nature 

conservation at Eden, New South Wales: A review of 

theory and practice. In D.A. Saunders, G.W. 

Arnold, A.A. Burbidge & A.J.M. Hopkins (eds) 

Nature Conservation: The role of remnants of native 

vegetation. (Surrey Beatty and Sons Pty Ltd: 

Chipping Norton). 

Richards B.N., Bridges R., Curtin R., Nix H.A., 

Sheppard R. & Turner J. (1990) Biological diversity of 

the South East Forests (Australian Government 

Publishing Service: Canberra). 

Riddell S. (1997) Willows in the far south coast 

streams of NSW. Report prepared for Far South 

Coast Willow Task Force and Department of Land 


Ryan D.G., Ryan J.E. & Starr B.J. (1995) The 

Australian Landscape – observations of explorers and 

early settlers. Murrumbidgee Catchment 

Management Committee, Wagga Wagga, NSW. 

Semple W.S. (1997) Native and naturalised shrubs 

of the Bathurst Granites: Past and present. 

Cunninghamia 5(1): 49–80. 

Sinai D.L. (1995) An assessment of bedload 

sediment delivery in an actively eroding catchment: 

Wolumla Creek Catchment, South Coast NSW. 

Honours thesis, School of Earth Sciences, Macquarie 

University. 

South East Forests Koala Research Project (1998) 

Modelling Koala Habitat and Use in Forests of the 

Eden Region. Report commissioned by RACAC for 

the Eden Comprehensive Regional Assessment. 

South East Forests Conservation Council, Bega. 

SFNSW (1998) A Field Guide to the South East 

Forests of New South Wales, vols 1–2. State Forests of 

New South Wales, Sydney. 

SFNSW (1994) Proposed forestry operations in Eden 

Management Area. Environmental Impact Statement. 

State Forests of New South Wales, Sydney. 

Stahl N. (1997) Stream salinity in Bega catchment, 

NSW: Development of a conceptual model and 

rapid appraisal technique. Honours thesis, 

Department of Forestry, ANU. 

Stone C. (1996) The role of psyllids (Hemiptera: 

Psyllidae) and bell miners (Manorina melanophrys) in 

canopy dieback of Sydney blue gum (Eucalyptus 

saligna sm.) Australian Journal of Ecology 

21: 450–458. 

Stone C. (1999) Assessment and monitoring of 

decline and dieback of forest eucalypts in relation to 

sustainable forest management: A review with case 

study. Australian Forestry (in press). 

Thoms M. & Bergs M. (1994) The character and 

sources of sediment to Pambula Lake. Interim 

report. A report prepared for the NSW Environment 

Protection Authority, Wollongong. 

Tulau M.J. (1997) Soil Landscapes of the Bega-Goalen 

Point 1:100 000 Sheet. Soil Conservation Service of 

NSW Soil Landscape Series. Department of Land 


Turner J., Lambert M. & Dawson J. (1996) Water 

Quality Monitoring Strategies for Forest Management: 

A Case Study on the Towamba Valley Catchment. 

Research Paper No. 31, Forest Research and 

Development Division, State Forests of New South 

Wales, Sydney. 

Turner J., Lambert M & Dawson J. (1997) Bega 

Valley Catchments Water Quality Monitoring Study. 

Report to Far South Coast Catchment Management 

Committee. 

Webb, McKeown and Associates (1997) Merimbula 

and Back Lakes Estuary Management Study and 

Management Plan. Report to Bega Valley Shire 

Council. 

Whelan R.J. & Hibberd, J.K. (1992) Mammals and 

habitat in the Bega valley: The importance of remnant 

forest on private property. National Estate Grants 

Scheme. Department of Planning, Sydney. 

97

Appendices


Appendix 1 

Vegetation types described for 


A brief description of the relevant vegetation types is 

provided below. Each vegetation type has been given 

a descriptive name which refers to the vegetation 

structure, nature of the understorey (grassy or 

shrubby), or a particular or general location where it 

occurs. Each type also has a number, 1 to 70, and 

W1 to W6 for types found only in Wadbilliga 

National Park (which was added late to the study 

area). The information in this section summarises 

that in Keith and Bedward (1999), which provides a 

fuller written description of each vegetation type, 

diagnostic species and distribution. This information 

is presented only in the form of tables in Keith and 

Bedward (1998). In some instances indicator species 

for the various vegetation types are those that have 

been found to be useful in the field during this study, 

rather than those given by Keith and Bedward 

(1998, 1999) or Keith, Bedward and Smith (1995), 

or Keith (1994). 

It should be noted that where vegetation types are 

described as being ‘mapped’ in particular areas, this 

is a shorthand way of saying that the vegetation 

model predicts that they will occur there. The 

production of the map derived from the vegetation 

model is described in sections 2.1 to 2.6. With 

regard to specific locations, if a vegetation type is 

known to occur there from ground surveys it is 

generally described in the following section as 

‘occurring’ there. If a vegetation type is predicted by 

the model but has not been verified on the ground, 

it is generally described as ‘mapped’ for that area. 

Estuarine vegetation types 

There are four seagrass communities, which occur 

below low-tide level in coastal saline lakes and 

estuaries. The most common is Seagrass Meadows 

(Zostera), 70, dominated by Zostera capricorni, 

which is found in most lagoons and estuaries in the 

area. The others are Seagrass Meadows 

(Halophila), 67, dominated by Halophila ovalis, 

Seagrass Meadows (Posidonia), 68, dominated by 

100 

Posidonia australis and Seagrass Meadows 

(Ruppia), 69, dominated by Ruppia polycarpa and 

R. megacarpa. All these types may occur in single or 

multiple species stands. Type 67 occurs mostly in 

Wallagoot, Cuttagee and Wallaga Lakes; 68 mostly 

in Merimbula Lake and the Bermagui River; and 69 

in Lake Curalo, Middle Lagoon in Mimosa Rocks 

National Park, Baragoot Lake and Wallaga Lake. 

Two mangrove communities fringe some coastal 

lakes and estuaries, occurring in the lower tidal zone. 

The most widespread is Estuarine Wetland (Grey 

Mangrove), 66, which consists generally of single 

species stands of the grey mangrove, Avicennia 

marina. Old stands consist of trees to about four 

metres in height, while younger stands, or those in 

more marginal habitat further from the water’s edge, 

may be stunted shrubs. Very much less common is 

Estuarine Wetland (River Mangrove), 65, 

dominated by Aegiceras corniculatum. This small tree 

or shrub mangrove prefers less saline conditions than 

Avicennia and is usually found further upstream in 

estuaries. It is known from only two sites in the 

Shire, Bermagui River (where there are quite well 

developed stands) and Merimbula Lake. The species 

is at the southern limit of its distribution at 

Merimbula Lake. 

On the landward side of the mangrove belt there is 

usually a strip of variable width of Salt Marsh, 64. 

This occurs on mudflats which receive regular or 

occasional tidal inundation. On the seaward side the 

dominant plant is the succulent herb, glasswort, 

Sarcocornia quinqueflora, with grey mangrove often 

present as a shrub. Another shrub, Sclerostegia 

arbuscula is also common. On the landward side, 

where tidal influence is less frequent, are grass 

(Sporobolus virginicus) and sedge zones (Juncus 

kraussii, Baumea juncea), with patches of common 

reed (Phragmites australis) sometimes present. 

Salt marsh may grade into another variable width 

band of Estuarine Wetland Scrub, 63. The 

dominant species here is generally swamp paperbark, 

Melaleuca ericifolia, with the small tree or shrub 

Myoporum acuminatum sometimes present. Swamp 

oak-dominated stands occur in the Bermagui estuary


and on Wallaga Lake, but this is the southern limit 

for this species. Occasionally, possibly on more 

steeply sloping shores where inundation is very 

uncommon, Melaleuca armillaris replaces 

M. ericifolia. 

Beach strand, dune and headland 


On ocean beaches Beach Strand Grassland, 62, 

occurs. The dominant species is spinifex grass, 

Spinifex sericeus, with a limited range of other 

salt-tolerant grasses, herbs and sedges potentially 

occurring. This community grades into Coastal 

Scrub, 61, which starts as wind-pruned prostrate 

coast wattle (Acacia sophorae) at the back of the 

beach and grades into taller forest dominated by the 

small tree Banksia integrifolia, with a shrubby 

understorey including Acacia sophorae, Leucopogon 

parviflorus and Monotoca elliptica. Ground cover 

generally includes bracken (Pteridium esculentum) 

and Lomandra longifolia. This combination of 

species occurs on sand, often on dunes behind 

beaches, but also on headlands. Where sedimentary 

rocks rather than sand are the substrate another 

species assemblage usually occurs which includes the 

trees Allocasuarina verticillata and Melaleuca 

armillaris and shrubs Westringia fruticosa, Alyxia 

buxifolia and Zieria littoralis. Where exposure to 

salt-laden winds is extreme the tree species are often 

dwarfed to shrubs. This latter species assemblage is 

typical of rocky headlands, though there is some 

species overlap between dune and headland sites. 

Rock scrubs 

There are four vegetation types described which 

grow in skeletal soils on rock outcrops or steep rocky 

slopes and cliffs (in addition to that described for 

rocky headlands above). 

Rhyolite Rock Scrub, 51, is restricted to rhyolite 

outcrops west of Pambula. Typical species are the 

small tree or shrub Melaleuca armillaris, the shrub 

Kunzea ambigua, and a number of other shrubs 

which are endemic to these outcrops. These include 

a number of threatened species: Phebalium ralstonii, 

Westringia davidii, Zieria parrisiae and Zieria 

buxijugum. Although there are other areas of rhyolite 

outcropping in the Shire, the rhyolite endemics seem 

to be largely restricted to the area west of Pambula. 

An exception is the Narrabarba Hill Flora Reserve 

west of Nadgee, which carries the only known 

populations of Acacia constablei. 

Mountain Rock Scrub, 52, occurs on granitoid 

outcrops at 500–1000 metres elevation in the 

south-west of the Shire (for example, Mount Poole, 

White Rock Mountain). Typical species include the 

shrubs Kunzea ambigua, Hakea macraeana, Boronia 

anemonifolia and Hovea purpurea, and the herbs 

Bulbine semibarbata and Calandrinia calyptrata. 

Rocky Tops Dry Shrub Forest, 3, sometimes has a 

canopy of eucalypt trees, but because this may be 

quite sparse, and because it grows in similar 

situations to the rock scrubs, it is included here with 

this group. Typical species are the gully gum 

(Eucalyptus smithii) with various other eucalypts 

occurring less commonly, and a range of shrubs 

including, on drier sites, Beyeria lasiocarpa, Cassinia 

longifolia, Haloragodendron bauerlenii and Olearia 

iodochroa and, on wetter sites, the trees Pittosporum 

undulatum and Ficus rubiginosa. Orchids, 

Dendrobium speciosum and D. striolatum may grow 

on sheltered rock faces. This vegetation type may 

occur on Ordovician or Devonian metasediments or 

granitoids and is scattered in the southern part of the 

Shire (for example, Burragate Peak, Big Jack 

Mountain, Mount Imlay). 

Acacia Scrub, 4, has a closed canopy formed by a 

dense, apparently even-aged growth of wattles, 

usually Acacia silvestris but occasionally, in 

Wadbilliga National Park, A. blayana, and occurs on 

steep rocky slopes on Ordovician metasediments. 

The understorey is usually very sparse, and includes 

the tall shrubs Beyeria lasiocarpa, Pomaderris 

brogoensis and Eriostemon trachyphyllus. Most 

occurrences are within Wadbilliga National Park and 

South East Forests National Park (Bemboka 

Section), but large stands occur on private and 

leasehold land in the Brogo Pass, and on private 

property on Alsops Creek nearby. Dense stands of 

other wattle species resulting from clearing or 

wildfire, generally Acacia mearnsii in the coastal 

valleys and A. dealbata on top of the escarpment, do 

not belong to this vegetation type as it has been 

defined by Keith and Bedward (1998). The latter 

species may occur on any geology and are not 

confined to steep slopes. 

101


These four vegetation types possibly do not fully 

describe the range of rock scrub assemblages present 

in the Shire. There are extensive areas of granitoid 

outcropping on Little Mumbulla Mountain and in 

Wadbilliga National Park, and of metasediments in 

South East Forests National Park (Bemboka 

Section), which were not sampled but were mapped 

as 52. Surveys of these could result in the creation of 

additional rock scrub types, or the broadening of the 

list of indicator species for some of the above types 

(see Appendix 3, p. 122). 

Heaths 

Heaths are generally treeless or very sparsely treed 

vegetation types found on shallow infertile soils or 

deeper infertile sand deposits. In the Bega Valley 

Shire they are largely confined to the area south of 

Eden, where Devonian sandstones and Tertiary 

sediments are more common. An exception is 

Montane Heath, 53, dominated by Allocasuarina 

nana. This occurs largely outside the Shire boundary 

along the western edge of Wadbilliga National Park, 

and around Bombala. 

Mount Nadgee Heath, 54, is restricted to rocky 

Devonian sandstone plateaux around Mount 

Nadgee and near Green Cape. It consists of a diverse 

range of sclerophyllous shrubs, with a dense ground 

cover of sedges and herbs. Occasional emergent 

brown stringybark (E. baxteri) are found here at 

their northern limit of distribution. 

Coastal Lowland Heath, 55, is found on flattish 

terrain on coastal Tertiary alluvium and recent sands 

below 100 metres elevation, mostly in Ben Boyd 

National Park and Nadgee Nature Reserve, south of 

Eden. Typical species are Allocasuarina paludosa, 

Banksia paludosa, Hakea spp, with a wide range of 

other shrubs and herbs. There are some very small 

and isolated patches of this vegetation type north of 

Eden (for example, north of Gillards Beach in 

Mimosa Rocks National Park, and on the 

Kangarutha Track in Bournda National Park.) 

Hinterland Heath, 56, occurs in the granitoid 

hinterland south-west of Eden, mostly below 

500 metres elevation. It is associated with small 

seepage areas on hill slopes. Individual patches are 

generally quite small, but the type is widely scattered 

within Yambulla and Timbillica State Forests, west 

of the Princes Highway. 

102 

Lowland Swamp, 57, overlaps in both distribution 

and species complement with 56, but is found in 

broad flat drainage lines with permanently 

waterlogged soils. It occurs on several geologies, west 

from Nadgee. Large stands are reserved in Sidling 

Swamp Flora Reserve, west of the Princes Highway. 

Typical species include Melaleuca squarrosa, Epacris 

paludosa, Sprengelia incarnata, Xyris spp., with 

fringing trees Eucalyptus ovata and E. conspicua. 

Sedges may dominate in the most low-lying areas. 

Scattered occurrences are also mapped for more 

northerly parts of the Shire, though these are likely 

to have a rather different species composition. 

Riparian and freshwater swamp 


This is a diverse group including shrub and 

tree-dominated types, occurring in river beds and on 

banks, in boggy drainage lines and in permanent and 

semi-permanent wetlands. 

Southern Riparian Scrub, 38, occurs in rocky river 

beds and on lower banks, principally south of 

Pambula (for example, Yowaka, Wallagaraugh and 

parts of Towamba rivers), on granitoids, Devonian 

sediments and rhyolite. This vegetation type shares a 

core group of species with Northern Riparian Scrub 

(Acacia floribunda, Callistemon subulatus, Lomatia 

myricoides, Leptospermum emarginatum) but is 

distinguished by also including Melaleuca armillaris, 

Calytrix tetragona, Grevillea linearifolia and 

Leptospermum scoparium. 

Northern Riparian Scrub, 39, occurs in the same 

situations, but is confined to granitoids in the rivers 

draining the drier rain shadow valleys, namely 

Tantawangalo and Candelo Creeks, Bemboka River 

and parts of the Brogo River. Distinctive species in 

these rivers include Melaleuca parvistaminea, Hakea 

microcarpa and Acacia elongata. Most other species 

are shared with 38. 

Riverine Forest, 40, is dominated by river oak, 

Casuarina cunninghamiana, with understorey species 

including the small tree Acacia mearnsii, shrubs 

Acacia floribunda and Hymenanthera dentata and 

grasses Microlaena stipoides and Oplismenus aemulus. 

It was originally mapped as occurring on river banks 

only on the lower Brogo River but Version 4 of the 

map includes occurrences on Narira Creek, Murrah 

River and tributaries of the Brogo River. However,


Casuarina cunninghamiana reaches its southern limit 

of distribution at about Bega, so 40 does not occur 

south of Bega. 

Swamp Forest, 58, is found along small drainage 

lines at 300–650 metres elevation in the southern 

hinterland on granitoids. Individual stands may be 

small but occurrences are widespread. There may 

also be small unmapped occurrences in the 

Wyndham area. Soils are permanently moist to 

waterlogged, with standing water occurring 

occasionally. Canopy species are swamp gum 

(E. ovata) and ribbon or manna gum (E. viminalis) 

with Acacia melanoxylon sometimes present. 

Understorey consists of a dense ground cover of 

Lomandra longifolia, or various sedges (Carex 

appressa, Cyperus lucidus) and herbs. 

Subalpine Bog, 59, occurs in waterlogged soils in 

broad flat valleys on the headwaters of mostly 

west-flowing streams at high elevations. Most 

occurrences are outside Bega Valley Shire, since the 

Shire’s western boundary follows the watershed 

along the top of the escarpment. However there are a 

few occurrences within the Shire, most notably Bega 

Swamp in Wadbilliga National Park. There is also an 

occurrence mapped for the headwaters of the 

Towamba River, south-east of Cathcart. Type 59 

consists of a mixture of wet heath (Baeckea utilis, 

Epacris paludosa, Hakea microcarpa) and wet 

grassland (Poa costiniana and numerous herbs) with 

various sedges (Empodisma minus, Restio australis). 

Floodplains Wetland, 60, is found at low 

elevations. Most mapped occurrences are located on 

the floodplains of major rivers (Murrah, Bega, 

Towamba) or along smaller drainage lines flowing 

into coastal lagoons. The former sites are arguably 

the most severely affected by agriculture in the 

region, so remnants of the former native vegetation 

are few and often severely degraded. Type 60 is 

recognised as including quite a range of species 

assemblages, ranging from Melaleuca ericifolia closed 

scrub, to reed beds dominated by Phragmites australis 

or other reed and sedge species, to permanent 

lagoons with emergent and submerged aquatic 

plants (see Appendix 3, p. 122). One occurrence 

which has been recognised but originally not 

mapped is in lower order drainage lines within the 

agricultural areas of the Shire. Many of these have 

been subject to draining and pasture improvement, 

or have been drained by gully erosion since 

European settlement, but a few still carry reed beds 

or other wetland vegetation. Some known 

occurrences have been included in Version 4 of the 

map (Keith & Bedward 1999). 

Rainforest vegetation types 

Five rainforest vegetation types have been described 

which occur in Bega Valley Shire. A sixth, which 

occurs only at the northern end of the Shire, was not 

sampled (see Appendix 3, p. 122). 

Dry Rainforest, 1, is dominated by Port Jackson or 

rusty fig, Ficus rubiginosa, often with Pittosporum 

undulatum and sometimes Alectryon subcinereus in 

the canopy as well. Forest red gum (Eucalyptus 

tereticornis) and kurrajong (Brachychiton populneus) 

are common emergents. Understorey is very sparse 

and consists of ferns, herbs and grasses. Vines are 

common, with Celastrus australis being the most 

characteristic species, though many others also 

occur. Dry Rainforest generally occurs as small 

patches interspersed through eucalypt forest (often 

Brogo Wet Vine Forest, 18), and is generally 

associated with large granitoid bedrock outcrops, 

though sometimes found on Ordovician 

metasediments. The rock outcrops provide this 

fire-sensitive community with some protection from 

fires, which is essential since the preferred habitat 

seems to be upper north-facing slopes and gully 

heads. Mapped occurrences are mainly in South East 

Forests National Park (Coolangubra Section) and 

the Towamba valley, which is the southern limit of 

distribution for Ficus rubiginosa. However, 1 is 

scattered throughout the agricultural areas of the 

Shire, mainly along the western edge from 

Murrabrine to Myrtle Mountain. Most extensive 

occurrences are in the Brogo area and around 

Meringola Peak west of Bega. Some of these will be 

included in Version 4 of the map (Keith & Bedward 

1999). Similar vegetation occurs in Wadbilliga 

National Park, though the extent is unknown. 

Myanba Eucalypt-Fig Forest, 2, is a similar forest 

type mapped in a restricted area on steep granitoid 

slopes of Myanba Gorge and nearby areas in South 

East Forests National Park (Coolangubra Section) at 

about 550 metres elevation. Various eucalypts occur 

in this type over a shrubby understorey including 

numerous Dry Rainforest elements such as Ficus 

103


rubiginosa, Hymenanthera dentata, the vine Celastrus 

australis and the rock orchid Dendrobium speciosum. 

Bunga Head Rainforest, 5, has a very restricted 

distribution. This is a littoral rainforest type, 

occurring very close to the sea. Only one occurrence 

is mapped, at Aragunnu Beach in Mimosa Rocks 

National Park, but there may be another at Boulder 

Bay in Bournda National Park. Type 5 occurs in 

small gullies draining directly to the ocean. Principal 

tree species are Ficus rubiginosa, lilly pilly (Acmena 

smithii) and Pittosporum undulatum, which can be 

severely salt-pruned. Beneath the canopy vines and 

ferns are the most common understorey plants. 

Coastal Warm Temperate Rainforest, 6, is 

restricted to steep sheltered gullies, usually south- to 

east-facing, on metasediments mainly in the coastal 

range but extending well inland into Wadbilliga 

National Park, below 300 metres elevation. Main 

tree species are Acmena smithii, Pittosporum 

undulatum and Doryphora sassafras, with pencil cedar 

(Polyscias murrayi) also common. The understorey 

usually consists mainly of ferns, and numerous vines 

scramble over the canopy. A variant dominated by 

grey myrtle (Backhousia myrtifolia) has been 

included within 6, but should probably be described 

as a separate type (see Appendix 3, p. 122). 

Hinterland Warm Temperate Rainforest, 7, occurs 

in similar locations but at higher elevations than 6. 

There is considerable species overlap between the 

two types, with only a handful of species being 

distinctive to 7. The presence of Acacia melanoxylon, 

Hedycarya angustifolia, Eucryphia moorei, Olearia 

argophylla, Coprosma quadrifida or Pomaderris aspera 

helps to distinguish 7 from 6. 

Cool Temperate Rainforest, 8, occurs at higher 

altitudes again, generally over 700 metres elevation 

on the escarpment and on outlying mountains such 

as Burragate Peak and Mount Imlay. The most 

common dominant tree is black oliveberry 

(Elaeocarpus holopetalus), with blackwood (Acacia 

melanoxylon) and southern sassafras (Atherosperma 

moschatum) also occurring. There may be emergent 

brownbarrel (Eucalyptus fastigata) and the 

understorey is very sparse and includes soft tree fern 

(Dicksonia antarctica) and other ferns, some of 

which grow as epiphytes on tree fern trunks. Fieldia 

australis is another common epiphyte. Vines are 

104 

uncommon. Occasional stands, such as that on 

Burragate Peak, have pinkwood (Eucryphia moorei) 

as the canopy species. 

Eucalypt forests 

The 31 vegetation types discussed above all have 

very specific habitat requirements, and with the 

exception of some of the heaths, tend to be found in 

small, naturally fragmented stands. The eucalypt 

forests, on the other hand, tend to occur over much 

larger areas, though some are more restricted. Each 

has a preference for particular soil types, moisture 

regimes and degree of site exposure, which dictates 

how the types occur in a mosaic in the landscape. 

The eucalypt forest types are divided here into wetter 

and drier types. The wetter types are generally found 

on the upper escarpment, or in gullies and on 

sheltered slopes of the coast and hinterland. The 

drier types can be divided into dry shrub forests and 

dry grass forests. The former occur on the more 

infertile soils derived from metasediments, on ridges 

and exposed slopes throughout the Shire. The dry 

grass forests tend to be found on soils derived from 

granitoid rocks in the hinterland, often on 

undulating rather than steep terrain. ‘Wetter types’ 

should not be taken to be synonymous with wet 

sclerophyll forest, which refers to tall forests found in 

areas of high rainfall, generally with tall dense 

shrubby understorey. Some of the escarpment forests 

fall into this category, but most of the forests in the 

area fall somewhere between typical wet and dry 

sclerophyll, so that these terms are not particularly 

useful in describing the vegetation of the area. 

High altitude wetter eucalypt forest 

types 

High Mountain Wet Layered Forest, 9, is scattered 

along the top of the escarpment above 850 metres 

elevation, mostly on granitoid geology. It is a tall 

forest, often over 40 metres in height, with a tall 

dense understorey of small trees such as Acacia 

dealbata, Olearia argophylla and Pomaderris aspera. 

Ferns dominate the sparse ground cover, and tree 

ferns are common. Principal canopy species are 

brownbarrel (E. fastigata) and shining gum 

(E. nitens). 

Mountain Wet Layered Forest, 10, is similar, but 

tends to consist of pure stands of E. fastigata with


monkey gum (E. cypellocarpa) occasional. The tall 

understorey is dominated by Bedfordia arborescens, 

Olearia argophylla, Pomaderris aspera and tree ferns, 

with a sparse ground cover of ferns and scattered 

herbs. Type 10 occurs in similar locations to 9, 

mostly on granitoids, along the top of the 

escarpment south from Wadbilliga National Park. It 

is more common in the northern part of the Shire, 

since a steep escarpment does not occur in the far 

south. Outlying mountains such as Egan Peaks may 

also carry some 10. 

Tantawangalo Wet Shrub Forest, 11, is restricted to 

the upper Tantawangalo Creek catchment on 

moderate granitoid slopes at 800–1000 metres 

elevation. Tall trees, Eucalyptus fastigata, 

E. cypellocarpa and messmate (E. obliqua) form the 

canopy, with the shrubby understorey being more 

diverse than in 9 or 10. Typical shrubs include 

Tasmannia lanceolata, Bedfordia arborescens, 

Pomaderris aspera, Hakea eriantha and Coprosma 

quadrifida. Ground cover is relatively dense and 

includes Gahnia sieberiana and Dianella tasmanica. 

Mountain Wet Fern Forest, 12, occurs on steep 

sheltered granitoid slopes at 400–900 metres 

elevation on the escarpment slopes and outlying 

mountains (for example, Egan Peaks, Mount Poole). 

It is similar to 10, but differences include a wider 

range of tree species, more frequent occurrence of 

E. cypellocarpa, and a shorter and less continuous 

shrub layer with a greater diversity of species. Ferns, 

grasses and herbs form the fairly continuous ground 

cover. 

Mountain Wet Herb Forest, 15, is widespread on 

moist sheltered granitoid slopes above 500 metres 

elevation on the lower tableland range, south from 

Tantawangalo Creek. Principal species are 

E. obliqua, E. cypellocarpa, shrubs Leucopogon 

lanceolatus and Coprosma quadrifida with the ground 

cover including the grass Hierochloe rariflora and 

numerous herbs. 

Basalt Wet Herb Forest, 16, is dominated by 

E. fastigata, with E. cypellocarpa, E. viminalis, 

E. radiata and the smaller tree Acacia melanoxylon 

potentially also present. Shrubs are uncommon, and 

the understorey is dominated by a diverse range of 

herbs, growing among small tussock grasses with 

bracken and Lomandra longifolia. It is found on 

gentle slopes at the top of the escarpment on basalt 

or granitoids, north from about Cathcart. 

Flats Wet Herb Forest, 17, occurs in broad flat 

valleys around small drainage lines on the top of the 

escarpment and to the east and west, at elevations of 

300–1000 metres. It generally occurs in small stands 

because of its restriction to drainage lines and 

surrounding flats. About half of the predicted 

occurrence of this restricted vegetation type is in 

scattered remnants in agricultural areas around 

Wyndham, but these have not been verified. Typical 

species are the trees E. viminalis, sometimes with 

E. elata or E. radiata, the fern Blechnum nudum and 

a range of herbs. 

Kydra Flats Grass Forest, W4, occurs on flats and 

sheltered slopes on metasediments at 

1000–1200 metres elevation along the western edge 

of Wadbilliga National Park. Typical trees include 

E. viminalis, E. dalrympleana, E. radiata and 

E. fraxinoides, shrubs Leucopogon lanceolatus, 

Lomatia myricoides and Persoonia silvatica, and there 

is an almost continuous ground cover of grasses and 

herbs. 

Wetter eucalypt forest types of the 

coast and hinterland 

Hinterland Wet Fern Forest, 13, is widespread in 

gullies and on sheltered slopes from sea level to 

800 metres elevation, both on granitoids of the 

escarpment and metasediments on the coastal range. 

It tends to be replaced by 12 or 15 at higher 

altitudes. The main canopy species is E. cypellocarpa, 

often with yellow stringybark (E. muelleriana). The 

ferns Calochlaena dubia and Blechnum cartilagineum 

are diagnostic. Mesophyll (soft leaved) shrubs such 

as Pomaderris aspera and Elaeocarpus reticulatus are 

typical, and vines, particularly Smilax australis and 

Tylophora barbata are common. 

Hinterland Wet Shrub Forest, 14, is also 

widespread in gullies and on sheltered slopes, below 

about 500 metres elevation, predominantly on the 

metasediments of the coastal range. There is 

considerable species overlap with 13, but the ground 

cover is usually extremely sparse and the shrub and 

small tree layer quite dense. The wattles Acacia 

cognata and A. subporosa commonly occur, along 

105


with numerous vines and tangles of wire grass, 

Tetrarrhena juncea. 

Coastal Gully Shrub Forest, 34, is widespread in 

gullies and on sheltered slopes on metasediments of 

the coastal range below about 200 metres elevation. 

It commonly co-occurs with Coastal Foothills Dry 

Shrub Forest, 32. No eucalypt species is recorded as 

being diagnostic, since quite a wide range of species 

can occur in this type. However, common species 

include E. longifolia, E. muelleriana and 

E. cypellocarpa. The mesophyll shrub understorey is 

usually shorter and more open than in 14, and 

typical species include Pittosporum undulatum, 

P. revolutum, Notelaea venosa, Hibbertia aspera and 

various Cassinia and Ozothamnus species. The 

predominantly grassy ground cover may be quite 

dense and diverse. 

Wadbilliga River Valley Forest, W6, occurs on 

sheltered lower slopes and river flats in the lower 

reaches of the Tuross and Wadbilliga rivers in 

Wadbilliga National Park. Trees include E. elata, 

E. viminalis and Angophora floribunda with 

Casuarina cunninghamiana on the river banks. There 

is a dense understorey of mesophyll shrubs and small 

trees, often with rainforest elements. Vines are 

common. 

Lowland Gully Shrub Forest, 37, is most common 

in the Nadgee–Timbillica area south of Eden, below 

about 100 metres elevation, occurring on gully flats 

on Tertiary alluvium or deep colluvial sandy soils 

derived from Ordovician or Devonian 

metasediments. Typical trees are Angophora 

floribunda, E. cypellocarpa and E. globoidea, with the 

small trees Allocasuarina littoralis and Acacia 

longifolia dominating the understorey. Bracken, 

various grasses and herbs form a tall ground cover. 

Eucalypt dry grass forests 

There is a group of four dry grass forests found 

almost solely in the agricultural areas of the northern 

rain shadow valleys (18–21) and another group of 

four found in the granitoid hinterland south-west of 

Eden (27–30). Another two (31 and 35) occur 

mostly on steeper country along the lower slopes of 

the escarpment, although extending into less steep 

western parts of the rain shadow valleys. 

106 

Brogo Wet Vine Forest, 18, occurs on hilly terrain, 

mostly on granitoids, but occasionally on small 

outcrops of Ordovician metasediments, in the wetter 

parts of the Bega valley and adjacent cleared valleys. 

Mapped as occurring from Murrabrine in the north 

to Meringola Peak west of Bega, it is also found 

scattered along the western edge of the Bega valley 

(for example, Myrtle Mountain, Tantawangalo, 

Bemboka). Typical trees are forest red gum 

(E. tereticornis), coast grey box (E. bosistoana), Acacia 

mearnsii, A. implexa, with rainforest elements Ficus 

rubiginosa, Pittosporum undulatum and Alectryon 

subcinereus. The latter trees, where continuous, form 

small patches of Dry Rainforest (1). Shrubs may be 

dense or patchy and include Cassinia trinerva, 

Hymenanthera dentata, Breynia oblongifolia and 

Abutilon oxycarpum. The ground cover consists of a 

wide range of grasses, ferns and herbs, or 

occasionally of a dense stand of hillside burrgrass, 

Cenchrus caliculatus. 

Bega Wet Shrub Forest, 19, occurs in gullies, on 

river flats and sheltered slopes below about 

250 metres elevation on granitoids, alluvium or 

Ordovician metasediments. Of the four types found 

primarily on agricultural lands, 19 is the one most 

likely to occur outside these areas, extending up river 

valleys into the hinterland to some extent. Within 

the dry rain shadow valleys, which have been largely 

cleared for agriculture, the main diagnostic species is 

blue box (E. baueriana) with E. elata also common. 

In other occurrences of 19, E. elata tends to be the 

main tree species, but it may be joined or replaced by 

E. viminalis, E. angophoroides or E. radiata on some 

sites. The understorey may consist of a dense shrub 

layer in which the main species are Cassinia trinerva, 

Hymenanthera dentata and sometimes Kunzea 

ericoides and Leptospermum brevipes. In other cases 

the shrub layer is sparse and a dense and diverse 

ground cover of grasses and herbs occurs. Bergalia 

tussock, Carex longebrachiata, is often present. 

Type 19 could also be placed with the wetter forest 

types described above, but is placed with Dry Grass 

Forests because of its similar distribution and land 

use pattern. 

Bega Dry Grass Forest, 20, is found on undulating 

or sometimes hilly terrain in the area between 

Cobargo in the north and Myrtle Mountain in the 

south. Small occurrences are mapped for the


Towamba valley, but may not be correct. Small 

occurrences are also mapped in some coastal 

locations where more fertile soils derived from 

plutonic rocks occur (Goalen Head, Tanja). 

Common tree species are E. tereticornis, E. globoidea 

and Angophora floribunda, with a patchy shrub/small 

tree layer typically containing Acacia mearnsii, 

Bursaria lasiophylla and Ozothamnus diosmifolius. 

There is a dense and diverse ground cover of grasses 

and herbs. Themeda australis and Microlaena stipoides 

are the most common grasses, but numerous others 

may be present. 

Candelo Dry Grass Forest, 21, occurs on 

undulating terrain in the driest parts of the Bega 

valley, with small patches also in the upper Towamba 

valley. Many species are common to 20 and 21, but 

those which help to distinguish 21 include yellow 

box (E. melliodora) and a number of daisy species 

(Calotis lappulacea, Chrysocephalum apiculatum, 

C. semipapposum, Brachycome ciliata). Most 

remaining stands are highly disturbed so it is 

difficult to determine what their original structure 

may have been like. Shrubs may be absent or quite 

dense, with blackthorn, Bursaria lasiophylla, the 

main native species. Exotic shrubs such as hawthorn 

(Crataegus monogyna), African boxthorn (Lycium 

ferocissimum) and briar rose (Rosa rubiginosa) often 

invade this vegetation type. The ground cover is a 

dense and diverse mixture of grasses and herbs, with 

Themeda australis usually the dominant grass species. 

Escarpment Dry Grass Forest, 35, occurs along the 

western edges of the Cobargo and Bega valleys, and 

all around the Towamba valley margins on moderate 

to steep slopes at elevations of 150–700 metres. It is 

found mostly on granitoids, but occasionally on 

basalt. Maiden’s gum (E. maidenii) dominates the 

canopy, with E. globoidea and E. bosistoana also 

present. Typical shrubs are Cassinia longifolia and 

Indigofera australis, and the ground cover may be 

dense or very sparse depending on site conditions. 

Typical species are the grass Danthonia longifolia and 

herbs Chenopodium carinatum and Plectranthus 

parviflorus. 

Hinterland Dry Grass Forest, 31, occurs 

extensively in the hinterland, south from Myrtle 

Mountain. It is found on undulating terrain at 

250–700 metres elevation, mostly on granitoids but 

also on Ordovician metasediments. Eucalyptus 

globoidea is the main tree present, with several other 

species commonly occurring (E. maidenii, 

E. angophoroides, E. cypellocarpa, E. sieberi). Shrubs 

are sparse, with the ground cover consisting 

predominantly of grasses, graminoids such as 

Lomandra longifolia, herbs and bracken. 

Waalimma Dry Grass Forest, 27, has a restricted 

distribution around Mount Waalimma, close to the 

Victorian border in the south-west hinterland. It 

occurs on ridges and upper slopes in undulating 

terrain on granitoids at 350–500 metres elevation. 

Typical trees include E. angophoroides, E. globoidea, 

E. tricarpa and E. bosistoana. A distinctive indicator 

species is red box, E. polyanthemos ssp. vestita. Shrubs 

are sclerophyllous and include Acacia mucronata, 

Epacris impressa, Lissanthe strigosa and Astroloma 

humifusum. There is a dense and diverse ground 

cover of grasses, herbs and graminoids such as 

Gahnia radula and Lomandra multiflora. 

Wog Wog Dry Grass Forest, 28, is found on gentle 

lower slopes on granitoids in the Wog Wog Creek 

area, mostly within South East Forests National Park 

(Genoa Section). Main tree species are 

E. agglomerata and E. maidenii. Shrubs are sparse, 

and the ground cover consists of a dense and diverse 

grass and herb assemblage. 

Nalbaugh Dry Grass Forest, 29, occurs on lower 

slopes on granitoids at 500–850 metres elevation at 

the southern end of the escarpment range. About 

half of its occurrences are outside Bega Valley Shire. 

Typical trees include E. globoidea, E. cypellocarpa, 

E. sp. aff. radiata and E. sieberi. Shrubs include 

Cassinia aculeata, Leucopogon lanceolatus and Epacris 

impressa. There is a diverse grass/herb ground cover. 

Wallagaraugh Dry Grass Forest, 30, occurs in 

small stands on lower slopes on a range of geologies 

between 100 and 500 metres elevation in the 

southern hinterland. It occurs in wetter sites than 

29, and may grade into 58, Swamp Forest, in 

drainage lines. Typical tree species are E. cypellocarpa, 

E. globoidea, E. radiata, E. viminalis and 

E. angophoroides, with the small tree Acacia mearnsii 

and the shrub Exocarpos strictus common. There is a 

diverse ground cover of grasses, Lomandra longifolia, 

herbs and bracken. 

107


Eucalypt dry shrub forests 

This large group of vegetation types tends to occur 

on the poorer soils and steeper terrain in the Shire. It 

can be divided into a low to moderate altitude group 

and a moderate to high altitude group, although 

some types occur over quite a wide altitudinal range. 

There are eight dry shrub forest types which are 

predominantly coastal to hinterland in occurrence, 

and 11 which occur mainly at higher elevations. 

Dune Dry Shrub Forest, 36, occurs only within a 

few hundred metres of the sea or coastal lagoons, on 

well-drained flats on Holocene sands. Such habitat is 

quite limited in Bega Valley Shire compared with 

other coastal areas. Small stands are scattered along 

the coast from Bermagui south to Boydtown. The 

dominant tree is generally bangalay, E. botryoides, 

although it may occasionally be replaced by other 

species. The understorey contains the small trees 

Banksia serrata or Banksia integrifolia, and shrubs 

including Monotoca elliptica and Acacia sophorae. 

Bracken and Lomandra longifolia are usually 

conspicuous in the ground cover, along with the 

grasses Themeda australis and Imperata cylindrica. 

Coastal Foothills Dry Shrub Forest, 32, is 

widespread though confined to Ordovician 

metasediments. It occurs on slopes and ridges at 

50–250 metres elevation on the lower parts of the 

coastal range, north of Eden. A large number of 

eucalypts occur in 32, but the most distinctive 

grouping is woollybutt (E. longifolia), coast grey box 

(E. bosistoana) and red ironbark (E. tricarpa). The 

distinctive dry spotted gum-ironbark forests 

(Corymbia maculata and E. paniculata or E. tricarpa) 

with shrubby understorey are considered part of 32 

since they occur in similar sites and share many 

understorey species. (Wetter spotted gum forests are 

included in 34.) The small tree Allocasuarina 

littoralis is often prominent in the understorey, and 

forms dense stands after disturbance such as logging 

or fire. Ground cover is sparse and includes the small 

shrub Platysace lanceolata and the large tussock grass 

Chionochloa pallida. 

Coastal Range Dry Shrub Forest, 33, is widespread 

on dry slopes of the coastal and hinterland ranges, 

mainly on metasediments at 100–600 metres 

elevation. Its distribution is more southerly (south of 

Bega) and more inland than 32. Typical trees are 

108 

yellow stringybark (E. muelleriana), E. cypellocarpa, 

E. sieberi and E. smithii. There is an understorey of 

tall shrubs including Acacia falciformis, Persoonia 

linearis and Cassinia longifolia. Ground cover is 

sparse. 

Timbillica Dry Shrub Forest, 46A, is abundant 

within a restricted area in the middle to lower 

reaches of the Wallagaraugh catchment in the 

southern hinterland. It occurs in undulating terrain 

on granitoids at 50–300 metres elevation, with the 

main occurrences being west of the Princes Highway. 

Typical tree species are yertchuk (E. consideniana), 

blue-leaved stringybark (E. agglomerata) and 

silvertop ash (E. sieberi). Allocasuarina littoralis is 

common, along with a diverse range of sclerophyll 

shrubs including Monotoca scoparia, Epacris impressa 

and Hibbertia empetrifolia. Typical ground cover 

species are Gahnia radula, Caustis flexuosa, 

Lepidosperma laterale and Patersonia glabrata. 

Lowland Dry Shrub Forest, 46B, is similar to 46A 

in its understorey species but tends to be dominated 

by different tree species, particularly red bloodwood 

(Corymbia gummifera), with E. sieberi, E. globoidea 

and Angophora floribunda also occurring. Some 

stands, particularly those on sands, also contain 

blackbutt, E. pilularis. It is a more widespread type 

than 46A, occurring on low ridges and drier slopes 

in the coastal foothills below about 150 metres 

elevation from Bermagui almost to the Victorian 

border. 

Eden Dry Shrub Forest, 47, occurs in the southern 

part of the Shire, on metasediments at 

90–800 metres elevation from Nadgee to Mount 

Imlay. Typical trees are E. sieberi, Angophora 

floribunda and Allocasuarina littoralis in the 

understorey, with an open shrub layer including 

Acacia terminalis, Epacris impressa and Pultenaea 

daphnoides. Bracken dominates the ground cover, 

with Gahnia radula, Gonocarpus teucrioides and 

Dianella caerulea also common. 

Mumbulla Dry Shrub Forest, 48, is found only on 

Bega tonalite, a granitoid rock type occurring on 

Mumbulla Mountain and Dr George Mountain. 

Type 48 occurs on ridges at 150–650 metres 

elevation. Eucalyptus sieberi is the main tree, 

occasionally with E. agglomerata. Allocasuarina 

littoralis is often present, along with a diverse


sclerophyllous shrub layer including Bossiaea 

obcordata, Persoonia linearis, Correa reflexa and 

Platysace lanceolata. Ground cover is sparse. 

Coastal Dry Shrub Forest, 49, is quite similar to 

48, though it more often includes E. agglomerata. It 

is more widespread, being found on metasediments 

at 100–500 metres (and occasionally to 900 metres 

in South East Forests National Park (Bemboka 

Section) and Wadbilliga National Park) on ridges 

and dry slopes, and occurring throughout the Shire. 

Foothills Dry Shrub Forest, 44, occurs on ridges 

and drier slopes at 500–900 metres elevation, usually 

on granitoids but sometimes on metasediments, 

mostly in the south-western corner of the Shire and 

beyond it to the west. Most is located in South East 

Forests National Park (Genoa Section). Eucalyptus 

sieberi is the main canopy species, sometimes with 

E. globoidea. The shrubby understorey is open, and 

includes Acacia falciformis, Persoonia linearis and 

Leucopogon lanceolatus. 

Mountain Dry Shrub Forest, 45, is also found in 

the south-western corner of the Shire at 

600–1000 metres elevation mostly on granitoids. 

Only about half of the occurrences are within the 

Shire, mostly within South East Forests National 

Park (Genoa and Coolangubra Sections). It is similar 

to 44, but the shrub assemblage is slightly different. 

Mountain Sandstone Shrub Forest, 43, occurs in 

the far south-western corner of the Shire, in Genoa 

National Park, and to the west of the Shire. It occurs 

mostly on Genoa sandstone around Nungatta 

Mountain at 400–900 metres elevation, with 

outlying occurrences at Nalbaugh Plateau and Big 

Jack Mountain on granitoids. Dominant trees are 

E. sieberi and E. globoidea with a dense sclerophyll 

shrub understorey. 

Tableland Dry Shrub Forest, 26, is also found in 

the south-western corner of the Shire, with about 

half of its mapped occurrences being outside the 

Shire. It occurs on metasediments and granitoids at 

300–850 metres elevation. Main tree species are 

E. globoidea, E. radiata and E. cypellocarpa, with a 

low, open sclerophyll shrub layer including Epacris 

impressa, Monotoca scoparia and Leucopogon 

lanceolatus. 

Genoa Dry Shrub Forest, 50, is similar to 44 and 

45, although E. agglomerata is the main tree species. 

It is found in the far south-western corner of the 

Shire, and most of its occurrence is outside the Shire 

boundary. It occurs on Genoa sandstone at 

300–740 metres elevation, with an outlying 

occurrence mapped on Mount Imlay. 

Inland Intermediate Shrub Forest, 42, is found on 

sheltered slopes on all geologies, at 150–650 metres 

elevation in the coastal ranges and hinterland south 

from Merimbula. Characteristic tree species are 

E. obliqua, E. sieberi and E. cypellocarpa, with a 

dense shrub layer including Acacia obtusifolia, 

Epacris impressa and Persoonia linearis. A tall ground 

cover includes bracken and wiregrass, Tetrarrhena 

juncea. 

Wadbilliga Dry Shrub Forest, W1, is similar to 48 

and 49, and occurs at elevations of 400–1000 metres 

on metasediments, on narrow ridges and upper 

slopes in steep terrain in Wadbilliga National Park 

and South East Forests National Park (Bemboka 

Section). Various eucalypts may dominate, including 

E. sieberi, E. globoidea, E. agglomerata and 

E. fraxinoides. Jilliga ash, E. stenostoma is endemic to 

Wadbilliga, Bemboka and Deua National Parks, and 

may also be a dominant species in W1. 

Wadbilliga Gorge Dry Forest, W5, is found on 

moderate to steep dry slopes at 200–500 metres 

elevation, on metasediments and granitoids in 

Wadbilliga National Park and South East Forests 

National Park (Bemboka Section), with some 

outlying patches mapped in Murrabrine and Murrah 

State Forests. Main trees are Angophora floribunda, 

E. globoidea and Allocasuarina littoralis, with a 

typically sparse sclerophyllous shrub understorey. 

Mountain Intermediate Shrub Forest, 41, occurs 

on steep upper sheltered slopes, generally above 

700 metres elevation on the escarpment and 

southern tableland range, but sometimes at lower 

elevations, as in western Nadgee Nature Reserve, and 

on outlying peaks (Burragate Peak, Mount Imlay, 

Letts Mountain). White ash (E. fraxinoides) is the 

most characteristic species, with E. sieberi also 

possibly present. It has an open shrub layer 

including Persoonia silvatica and Leucopogon 

lanceolatus and a tall ground cover comprising 

109


clumps of Lomandra longifolia, Dianella tasmanica, 

Hierochloe rariflora and bracken. 

Wadbilliga Range Shrub Forest, W2, is similar in 

structure and species composition to 41, but 

includes several shrub species which reach their 

southern limit of distribution at about Brown 

Mountain. It occurs on similar steep, elevated, 

sheltered slopes in Wadbilliga National Park, and to 

the north in similar deeply dissected terrain. 

Distinctive shrub species are Choretrum candollei, 

Oxylobium ellipticum and Acrotriche leucocarpa. 

110 

Wadbilliga Heath Forest, W3, is found on dry sites 

on Devonian metasediments on the Wadbilliga 

Plateau at 1100–1350 metres elevation. It occurs in 

more exposed sites than W2, and less exposed than 

Montane Heath, 53. Mallee eucalypts E. kybeanensis 

and E. latiuscula may occur, or the taller trees 

E. pauciflora and E. dalrympleana. A heathy shrub 

layer includes Hakea dactyloides, Allocasuarina nana, 

Banksia canei and Dillwynia sericea.


Appendix 2 

Detailed assessment of map 

accuracy for selected 


Assessment of map accuracy was concentrated on: 

• those vegetation types most closely associated 

with agricultural areas (18, 19, 20, 21, 39, 40 

and 60); 

• coastal vegetation types which are subject to 

pressure from residential development and 

recreational use (32, 34, 46B, 36, 60, 61, 63, 64 

and 66); and 

• rainforests (1, 6 and 7) which have restricted 

distributions and are vulnerable to fire, and 

hence to human management practices. 

Assessment of map accuracy was made subjectively, 

based on the presence of indicator species listed for 

the vegetation types (Keith & Bedward 1998, 1999; 

Keith, Bedward & Smith 1995) or, in some cases, on 

indicator species which have been found to be useful 

for some vegetation types which are not so listed. 

Areas where the mapping seemed likely to be wrong, 

in the light of local knowledge, were selected for site 

visits. Since time constraints did not permit the 

checking of all such areas, representative sites were 

chosen on the basis of ease of access. This means that 

the assessment has been biased towards finding 

errors rather than assessing overall accuracy, since 

mapping which appeared likely to be correct was 

generally not checked. 

Some sites were not visited, but were checked against 

data held from previous work done in the area. Such 

data includes: 

• raw data sheets for CRA data collection; 

• species lists made on private properties assessed 

for the Remnant Bushland Advisory Service; and 

• a considerable number of species lists for various 

locations on private and public lands within the 

Shire. 

1 Dry Rainforest 

This vegetation type was mapped entirely from aerial 

photo interpretation so, in theory, should be quite 

accurate. In fact, because it mostly occurs in very 

small stands, some of it has been overlooked. On the 

map version currently held by Council it is mapped 

as occurring only in the Towamba valley and in one 

location north-west of Bega. There are some 

relatively large stands (totalling 30 hectares) mapped 

for South East Forests National Park (Coolangubra 

Section), and the total pre-1750 area is estimated at 

47 hectares, giving a reservation status of 64 per cent 

reserved. Because of the amount which has been 

overlooked on private property, this would be an 

underestimate of area (although the size of stands in 

Coolangubra has also been exaggerated) and an 

overestimate of reservation status. It is likely that 

rather more has been cleared than is indicated 

(11 per cent), and that rather more still exists, 

mostly on private property in agricultural areas. 

There is a problem with definition of this vegetation 

type. In the Bega valley and to the north, it often 

consists of very small patches containing a few 

individuals of Port Jackson fig, Ficus rubiginosa, 

under which a limited range of other rainforest 

species grow, embedded within another forest type, 

often 18 (Brogo Wet Vine Forest). The problem 

arises over when to call the vegetation type 18 with 

an occasional fig, and when to say that there is Dry 

Rainforest present. This may be part of the reason 

why stands on private land have not been mapped as 

1. Similarly in the Coolangubra area the same 

problem of definition occurs between 1 (Dry 

Rainforest) and 2 (Myanba Eucalypt-Fig Forest) or 3 

(Rocky Tops Dry Shrub Forest), which may also 

include fig and pittosporum. The size of Dry 

Rainforest stands in Coolangubra appears to have 

been exaggerated through confusion in air photo 

interpretation with type 3, at least in some instances. 

Actual occurrences of Dry Rainforest which were not 

originally mapped are: 

• on private property in steeper terrain within the 

dry rain shadow valleys, for example, east of the 

111


former Murrabrine State Forest (now National 

Park) off Quinlans Road and Gilberts Road; 

• off Upper Brogo Road near the northern shore of 

Brogo Dam; 

• scattered along the north-facing escarpment 

which parallels Warrigal Range Road between 

about Dorrigo Road and the Brogo Dam; 

• in the steep metasedimentary belt west of Upper 

Cobargo Road, Brogo; 

• in the Meringola Peak area west of Bega; 

• at Tantawangalo on north- to northwest-facing 

slopes visible to the south-east of Tantawangalo 

Mountain Road; and 

• at the foot of Myrtle Mountain in the valley of 

Goldens Creek. 

The presence of relict figs in paddocks or on 

roadsides suggests that Dry Rainforest (or 18) may 

have formerly occurred south of Bemboka off 

Mogilla Road, at Numbugga, and at Bald Hills and 

Lochiel west of Pambula. 

Generally it could be expected to occur on 

north-facing upper slopes or in the heads of 

northwest- to northeast-facing gullies anywhere 

where slopes are fairly steep within the agricultural 

areas of the Shire. Most occurrences are on granite 

where there are large boulders outcropping, since 

these provide some protection from the effects of fire 

for the fire-sensitive rainforest species. However, it is 

also present on Ordovician metasiltstone, where this 

occurs as small screens and outcrops within the 

predominantly granitoid agricultural areas. Stands 

are likely to have always been small and naturally 

fragmented because of its site preferences. The extent 

of clearing may not have been great, but the 1952 

fire may have destroyed some stands and frequent 

burning in marginal agricultural lands may have 

prevented recolonisation of suitable sites. 

One occurrence is known in Wadbilliga National 

Park in a similar steep, rocky, north-facing gully 

head, but the type is likely to be uncommon there. 

No stands were detected by air photo interpretation, 

although some may have been misinterpreted as 

Acacia Scrub (4). 

112 

In summary, for Dry Rainforest, the CRA vegetation 

map underestimated the original extent of the type, 

failed to record a number of current locations, and 

overestimated the amount and proportion of the 

type in reserves. Recent amendments to the map 

(Version 4) have included the addition of some of 

the Dry Rainforest stands mentioned above. 

4 Acacia Scrub 

This vegetation type was also mapped directly from 

air photo interpretation. Strictly speaking, ‘Acacia 

Scrub’ as used by Keith and Bedward refers only to 

wattle scrubs in which the dominant species is Acacia 

silvestris or, rarely, Acacia blayana. These scrubs occur 

on steep sites on exposed slopes and gully heads on 

Ordovician metasediments. Most stands are in 

Wadbilliga National Park, but there are large stands 

on leasehold land in the Brogo Pass, and on private 

land one to two kilometres to the south-west on 

Alsops Creek. 

The air photo interpretation team have mapped all 

dense stands of wattle as Acacia Scrub, regardless of 

what the dominant wattle species might be, and this 

has been carried over into the earlier version of the 

CRA vegetation map. The majority of the vegetation 

so labelled would be black wattle (Acacia mearnsii) 

regeneration on marginal agricultural areas, with 

stands on top of the escarpment probably being 

Acacia dealbata fire or logging regrowth. This 

mistake has been corrected on Version 4. Correction 

of the error has resulted in all stands except the 

genuine A. silvestris/blayana scrub being redefined as 

whatever eucalypt-dominated vegetation type is 

predicted for that area. 

Acacia Scrub, 4, has been substantially overestimated 

as a result of all wattle regeneration being mislabelled 

as this type on the earlier version of the map. 

Genuine type 4 can be readily distinguished on 

aerial photos by having a finer texture than Acacia 

mearnsii regeneration. Outside of Wadbilliga 

National Park and South East Forests National Park 

(Bemboka Section), it is confined to the vicinity of 

Brogo Pass. The error has resulted in an incorrect 

estimate of the proportion of 4 that is reserved. In 

fact, almost 100 per cent would be in reserves.


6 and 7, Coastal and Hinterland 

Warm Temperate Rainforest 

These two vegetation types were mapped from air 

photo interpretation and discriminated from each 

other on the basis of altitude, temperature and 

rainfall regimes. They are mapped as being widely 

scattered throughout the district in small stands, 

largely confined to sheltered gullies. The two types 

can often be found mapped in the same gully, one 

above the other, where their predicted distributions 

overlap (for example, Murrabrine, Yurammie). It is 

easier to consider the two types together, since they 

are very similar in species composition, habitat 

requirements, and threats, and the proportion 

estimated to be reserved is similar (42 and 55 per 

cent respectively). 

The map shows numerous small rainforest patches 

on private property, with mostly 6 mapped in the 

Brogo area and mostly 7 around the eastern edge of 

Murrabrine State Forest, along the western edge of 

the Bega valley and in the Wyndham area. It is 

probable that a large proportion of these patches are 

actually black wattle scrub (Acacia mearnsii). Some 

were checked against known areas (for example, 

around Brogo and north of Myrtle Mountain) and 

were definitely wattle rather than rainforest. While 

rainforest certainly does occur in these areas, it is not 

as extensive as has been mapped. It does often 

happen in steeper agricultural areas that small 

patches of rainforest are left in gully heads, while the 

lower gully, being less steep, is cleared then 

subsequently regenerates to black wattle scrub. The 

two blend together on aerial photos and are 

indistinguishable, which may account for some 

errors. In other instances, patches of rainforest have 

been mapped as occurring on slopes rather than in 

gullies. At least one of these is actually Dry 

Rainforest, but the majority are probably black 

wattle scrub. Rainforest (6) has also been mapped as 

occurring on the banks of the Brogo River in the 

Brogo Pass which is not correct. While there is 

rainforest in several gullies draining into the river in 

this area, the banks carry a mixture of Riverine 

Forest (40) and Bega Wet Shrub Forest (19). 

The extent to which rainforest has been 

over-mapped in State Forests and reserves has not 

been checked. The problem may be corrected by 

discarding the CRA rainforest layer and reverting to 

an earlier, and more conservative, air photo 

interpretation layer for non-eucalypt vegetation 

types. It is difficult to say what effect the 

over-mapping of rainforest has had on the estimates 

of area in reserves. While mapping of numerous 

small patches on private property will have tended to 

reduce the proportion estimated to be in reserves, it 

may well have been over-mapped in reserves as well. 

While the extent of Warm Temperate Rainforest on 

private property may have been exaggerated by the 

CRA vegetation map, it should not be assumed that 

it does not occur there. There are some quite large 

and diverse rainforest stands on private property 

within marginal agricultural areas, some of which are 

known to provide habitat for rainforest-dependent 

fauna and to contain regionally significant 

occurrences of plant species at their southern limits 

of distribution. There are also good rainforest stands 

on private property along the coast. They can be 

expected to occur in steep sheltered gullies, usually 

draining south to east. These sites provide the best 

topographic protection from fire. 

Another error with the rainforest mapping is that 

stands of the undescribed dry rainforest type 

dominated by Backhousia myrtifolia have been 

misidentified as vegetation type 6. Such stands 

mainly occur from Bermagui to Mumbulla State 

Forest on the coastal strip and in the coast range 

(personal observation; Floyd 1982), and in 

Wadbilliga National Park (Crawford 1992; Floyd 

1982). This vegetation type was not sampled at all, 

so that further field work would be required to 

describe it and predict its distribution. 

38 and 39, Southern and Northern 


Riparian Scrub was mapped from air photo 

interpretation and the two types separated on the 

basis of rainfall. Although they have been called 

‘southern’ and ‘northern’, Southern Riparian Scrub 

(38) is actually mapped as occurring at both ends of 

the Shire, while Northern Riparian Scrub (39) is 

more central in its distribution. 

Type 38 (southern) is mapped for the Brogo River 

downstream from the Brogo Dam to about the 

Angledale Crossing as well as in the more southern 

113


streams (Yowaka River, Towamba River below the 

Wog Wog confluence, and the full length of Imlay 

Creek and the Wallagaraugh River). 

Type 39 (northern) is mapped for: 

• the Brogo River from the Angledale Crossing to 

the Bega River confluence; 

• Tantawangalo Creek and the Bemboka River 

from where they enter private property to their 

confluence; 

• a small section of Candelo Creek; and 

• the Towamba River from where it enters cleared 

land to the Wog Wog confluence. 

No riparian scrub is mapped for Narira Creek or its 

tributaries, Dry/Murrah River, or any of the Brogo 

tributaries other than the lowest one kilometre of 

Double Creek, nor for Candelo or Tantawangalo 

Creek, Bemboka or Brogo River above the point at 

which they enter cleared land. In fact all of these 

latter streams carry riparian scrub in at least some 

parts of their upper reaches, generally where they are 

wide enough so that overhanging trees do not 

entirely shade the bed. Riparian scrub is best 

developed where the streambed receives full sun. 

Whether riparian scrub is actually present on Narira 

Creek is not known. It is probably correct that Dry/ 

Murrah River does not carry any significant amount 

above Quaama, since the reaches upstream of cleared 

land carry rainforest and the reaches within cleared 

land are mostly denuded of native vegetation. 

However, downstream from Quaama there is 

reasonably diverse riparian scrub, which persists 

through at least some parts of the middle sections of 

the river, which are within Biamanga National Park 

or State Forest. 

The Brogo River carries riparian scrub upstream of 

the Brogo Dam. Most of its tributary creeks (House, 

Alsops, Double and Desert) are relatively 

undisturbed, even where they pass through cleared 

land. All carry some riparian scrub. 

The upper section of the Towamba River between 

Burragate and Towamba carries 38, not 39 as 

mapped. Downstream sections have not been 

surveyed but are probably much depleted of 

114 

vegetation by clearing, river widening and sediment 

deposition. 

The range of species assemblages present in riparian 

scrub in the Shire could be better described by 

having three riparian scrub types (see Appendix 3, 

p. 122). Their distribution would be: 

• ‘northern’ on Narira Creek (if there is any 

riparian scrub remaining here), Murrah and 

Brogo Rivers and tributaries; 

• ‘central’ on Tantawangalo Creek, Bemboka River 

and tributaries; and 

• ‘southern’ on everything south of the Bega valley. 

It is the ‘central’ type occurring on the streams of the 

most intensively cleared parts of the Bega valley 

which is most depleted and least reserved. The figure 

of five per cent in reserves may well be generous for 

this type. 

40 Riverine Forest 

This vegetation type, dominated by Casuarina 

cunninghamiana, was mapped solely from air photo 

interpretation. It was originally mapped as occurring 

only on the Brogo River, between the lower end of 

the Brogo Pass and Angledale Crossing. Outside this 

area the presence of forest on the river banks made it 

impossible to detect whether or not Casuarina is 

present. In fact 40 is very much more extensive than 

this on the Brogo, occurring from at least as far 

upstream as the confluence with Brogo Swamp 

Creek (though not as far up as the Dam), right down 

to the Bega confluence, though in this lower section 

it is patchy. It also occurs on the Bega River 

downstream from Bega although, again, it is patchy. 

It is present on the lower reaches of several of the 

Brogo tributaries: Brown’s Creek for 1.5 kilometres 

upstream of its confluence, House Creek for about 

two kilometres to Warrigal Range Road, and Double 

Creek for at least three kilometres. It is also present 

on Narira Creek from about the Barrabaroo 

confluence two kilometres east of Cobargo down to 

Wallaga Lake, and a short distance up some 

tributaries, though again its occurrence is patchy. 

The lower Murrah River also carries Casuarina 

cunninghamiana from about the Mumbulla Creek 

confluence down.


Some amendments have been made to the most 

recent CRA vegetation map in the light of the above 

information, and Version 4 will show the 

occurrences listed above. None of this mapping of 

additional stands alters the reservation status of this 

vegetation type in the Shire, which is nil. 

18 to 21, rain shadow valley grassy 

forests complex 

These four vegetation types account for the bulk of 

remnant vegetation within the agricultural areas of 

the Shire, occurring primarily on soils derived from 

granitoids. Bega Wet Shrub Forest, 19, has a 

somewhat wider distribution, occurring on alluvium 

along coastal creeks and on metasediments as well as 

granitoids. 

A general comment can be made about these types as 

a group before considering the accuracy of mapping 

for each one. From Table 3.1 in Keith and Bedward 

(1998) it is possible to calculate the proportion of 

estimated pre-1750 stands currently reserved for 

each vegetation type. All four rain shadow valley 

grassy forest types come up as having a very low level 

of representation in reserves, 10 per cent for 18 

being the best, with four per cent for 19 and a mere 

0.5 per cent for 20 and 21. In fact the extent of 

reservation is even worse than it is depicted, since 

the vegetation map is inaccurate in showing types 18 

and 19 penetrating some distance into reserves 

surrounding the valley. Very small areas of 20 and 21 

are also mapped for public lands on the valley 

margins. It is doubtful if very many of these are 

correct. What in fact often happens is the reverse, 

with non-rain shadow valley forest types extending a 

short distance into the steeper parts of the cleared 

lands. Several sites were checked, either physically or 

against species lists held for particular valley margin 

locations, and no evidence of vegetation types 18, 20 

or 21 extending off privately owned lands was 

found. 

18 Brogo Wet Vine Forest 

This vegetation type seems to have been 

substantially over-mapped. Part of the reason for this 

is failure to take into account the effect of aspect on 

vegetation type. It is mapped as occurring uniformly 

on all aspects on steeper country within the Bega 

and Cobargo–Quaama valleys, at elevations above 

about 180 metres. In fact it appears to occur 

predominantly on northerly aspects. 

The map shows it as being most extensive in the 

Brogo and Verona areas north to about Murrabrine, 

which is correct. However, within these areas it 

would be unlikely to be even half as abundant as the 

map suggests. Most of the Verona area appears to 

actually carry vegetation type 19. Other steep areas 

close to valley margins, such as Brogo, often carry 

35, Escarpment Dry Grass Forest, in areas which 

have been mapped as 18. However, a substantial 

scatter of 18 around Meringola Peak west of Bega 

appears to be correct. 

The mapping of large areas of 18 along the margins 

of Wadbilliga National Park and South East Forests 

National Park (Bemboka Section) is wrong. The 

valley of Desert Creek in Bemboka Section was 

checked and no evidence of 18 occurring there was 

found. Occurrences along the southern boundary of 

Murrabrine State Forest (now National Park), and 

the western edge of Biamanga National Park and 

Mumbulla State Forest were not checked and may or 

may not be correct. 

Small areas of 18 which do occur within the Bega 

valley have been overlooked through confining its 

mapping to granitoids. There are small patches on 

Ordovician metasediments at Brogo (west of Upper 

Cobargo Road) and Tantawangalo. Small patches of 

18, or relict individual trees of Port Jackson fig 

suggesting former presence of 18, occur around the 

western rim of the Bega valley, for example, south of 

Bemboka, at the base of Myrtle Mountain, south of 

Tantawangalo Mountain Road and off Slaters Lane 

south-west of Bimbaya. 

19 Bega Wet Shrub Forest 

Mapping of 19 shows it as being the most 

extensively occurring of the four rain shadow valley 

grassy forest types. This is correct in that it does 

extend into surrounding public lands along drainage 

lines, whereas 18, 20 and 21 do not. However, the 

extent of this occurrence off private property has 

been exaggerated. Correcting this would reduce the 

estimate of four per cent reservation of vegetation 

type 19. For example, a large area mapped at the 

confluence of Mumbulla Creek and the Murrah 

River in Biamanga National Park appears from 

115


species lists held for that area to be incorrect. While 

there are patches of 19 scattered on small alluvial 

flats along the Murrah River, they are unlikely to add 

up to the size of the area mapped. 

On New England Creek in Murrabrine State Forest 

there are extensive areas mapped, but records from 

this area show that 19 is mapped as occurring too far 

up drainage lines. In the upper reaches it is replaced 

by 13 (Hinterland Wet Fern Forest). This is a 

common pattern, with the same thing happening 

around Stanton Rock in the South East Forests 

National Park (Yowaka Section). The degree to 

which 19 is tied to drainage lines has been 

underestimated in steeper areas, with the result that 

it has been mapped as being probably twice as 

abundant as it really is in these areas. This would also 

apply to areas in Coolangubra State Forest and 

South East Forests National Park (Coolangubra 

Section). Occurrences mapped between Imlay Creek 

and the Princes Highway in Timbillica State Forest 

will be deleted from the latest version of the map 

(D Keith, pers. comm.). 

On private property, 19 is mapped as being most 

extensive in valley margin areas in the following 

locations: 

1. east of the Princes Highway between Cobargo 

and Quaama; 

2. west of the Princes Highway at Verona and 

Brogo; 

3. Greendale, between the Brogo River and 

Mumbulla Mountain; 

4. north-east of Little Brown Mountain, near 

Bemboka; 

5. along the western edge of Bournda Nature 

Reserve east of Wolumla, and south-west of 

Wolumla; 

6. along the middle reaches of the Pambula River; 

7. north-east of Wyndham, along Kingfisher Road; 

8. south-east of Wyndham, along the Burragate 

Road; 

9. in the upper Towamba valley between Burragate 

and New Buildings Road; 

116 

10. from Lochiel south along Back Creek and 

Nethercote Roads. 

Results of ground truthing in the private property 

mapped areas are as follows: 

1. Cobargo-Quaama: 19 does occur extensively in 

lower parts of the landscape, with 20 at higher 

elevations. Exact locations may vary from those 

mapped but the proportions appear correct. 

2. Verona and Brogo: 19 is extensive around 

Verona, possibly more extensive than mapped. It 

is also common at Brogo, but in at least some 

locations it has been over-mapped. The large 

patch running up Desert Creek into South East 

Forests National Park (Bemboka Section) does 

not exist, and the extent around the southern 

edge of Bemboka Section has probably been 

exaggerated. 

3. Greendale: This area was not checked. 

4. Little Brown Mountain area: 19 is present here as 

a narrow riparian strip and in gullies; not in a 

large solid block as mapped. 

5. Wolumla area: 19 is present, but there are large 

areas of other non-rain shadow valley vegetation 

types extending west from the steeper country in 

Bournda Nature Reserve, so the total area of 19 

is probably less than mapped. 

6. Pambula River valley: There is a lot of 19 

present, but it appears to be more closely tied to 

riverbanks than is suggested. 

7. Kingfisher Road: a large block of 19 is mapped 

on lower southwest-facing slopes. This area in 

fact consists of a mosaic of vegetation types on 

low ridges and in broad flat gullies. In some areas 

Eucalyptus elata is present, but other than this the 

vegetation shows little resemblance to 19. 

Affinities with types 15 (Mountain Wet Herb 

Forest) and 17 (Flats Wet Herb Forest), which 

are mapped as occurring at higher elevations in 

the Wyndham area, seem stronger. Some gullies 

contain 58, Swamp Forest. 

8. Burragate Road: The area between Myrtle Creek 

and Burragate Road contains one of the most 

substantial blocks of 19 in near-natural 

condition in the Shire. Even so, 19 has been


over-mapped here, as it is more closely tied to 

drainage lines than is depicted. 

9. Upper Towamba valley: Remnants do indicate 

that 19 was extensive here on valley flats, but 

most has been cleared. The type is over-mapped 

here. 

10. Lochiel-Nethercote: Remnants on roadsides and 

creek banks indicate 19 may have been common 

on lower lying sites, possibly with 18 on more 

elevated sites (although 18 is not mapped for the 

area). Most of these areas have been cleared. 

Remaining vegetation around the valley margins 

more closely resembles 13 (Hinterland Wet Fern 

Forest) or 14 (Hinterland Wet Shrub Forest). 

There are some areas where 19 occurs but has not 

been mapped. It is present along Mataganah Creek, 

where it can be seen from New Buildings Road. It is 

probably more extensive in the Yowrie valley than is 

mapped, as most remnants visible from the road 

appear to be 19. The Eden CRA study area excludes 

everything to the north of the Yowrie Road and the 

Princes Highway beyond Cobargo. Most of the 

remnant vegetation in the Wandella valley, which is 

not included in the map, is probably 19, and it could 

also be expected to co-occur with 20 in the 

undulating cleared country north of Cobargo. It is 

also more common on the coast than mapping 

indicates (for example, Moncks Creek flats west of 

Wallagoot Lake) and on the Bega River floodplain 

(for example, on Reedy Swamp Road and Wallagoot 

Lane). 

Mapping suggests that 19 is uncommon in the drier 

centre of the Bega valley, with only very small 

patches shown in the Meringola Peak area and 

around Bimbaya. These have not been checked, but 

the general shortage of 19 in the drier areas appears 

to be correct. 

In general, it appears that in some areas of low relief 

(such as Cobargo) 19 may be as extensive as 

mapped. However, in some hilly areas it has been 

substantially over-mapped through underestimating 

the effects of aspect (it is generally only extensive on 

sheltered slopes), and the degree to which it is tied to 

drainage lines. 

Vegetation type 19 is also quite heterogeneous, 

containing three sub-types (see Appendix 3, p. 122) 

which appear to be related to site quality. One of 

these sub-types would not be readily recognisable as 

19 from the diagnostic species list in Keith and 

Bedward (1998), but is not referrable to any other 

type. This is the type in which Eucalyptus 

angophoroides, E. viminalis and E. radiata are 

common. It appears to be most extensive in the 

Murrabrine and Wyndham areas. It may be that the 

two sub-types occurring on poorer sites are better 

reserved than the sub-type from better quality sites. 

This latter typically contains blue box (E. baueriana) 

and is the sub-type most commonly occurring on 

agricultural land. 

21 Candelo Dry Grass Forest 

Mapping of 21 has tended to concentrate it in the 

western end of the Bega valley, although there are 

isolated occurrences mapped elsewhere (for example, 

Jellat Jellat, and near Rocky Hall). Some occurrences 

seem improbably close to the valley margins, even 

running up river valleys into areas which are 

otherwise mapped as containing various valley 

margin vegetation types. Some of these locations 

were checked and found either to be more like 20 (if 

some distance from the edge of surrounding forests) 

or some other non-rain shadow valley vegetation 

type from adjacent steeper public lands (if they were 

right at the cleared land–forest interface). Checking 

was done east of Yankees Gap, south of Bemboka 

and along the Bemboka River downstream from the 

hydroelectricity-generating station. Only at the latter 

site was any evidence of 21 found and, in this 

instance, it extended at least one kilometre less far up 

the valley than mapped. 

The distribution of 21 in the Bega valley occurs 

mainly in the centre of the valley from Tantawangalo 

north to about Pollocks Flat Road north of 

Bemboka, and from the Princes Highway west to 

about Mogilla Road. It is unlikely to extend any 

closer than to within one kilometre of the valley 

margins. Within this central area it is not mapped as 

extensively as it should be. For example, stands on 

Kameruka which are definitely 21 are mapped as 20. 

At Tantawangalo, remnant 21 on the speedway 

reserve is mapped as such, while that around the 

Tantawangalo church is mapped as 20. 

Another problem with the mapping of 21 is that 

some occurrences have not been mapped because the 

117


tree cover has been deemed too sparse for the sites to 

be recorded as remnant vegetation from Landsat 

imagery. If the grass and herb understorey is in good 

condition then remnants with trees at woodland 

densities (10–30 per cent canopy cover) can be just 

as valuable as those with a denser tree cover. Good 

stands of 21 which have been omitted for this reason 

occur on Kameruka Lane and Coronation Lane on 

Kameruka, and along Slaters Lane opposite the 

Tantawangalo speedway. 

Some small remnant patches with a denser tree cover 

in the centre of the Bega valley have also not been 

mapped, and would probably be vegetation type 20 

or 21. Known examples are near the corner of 

Alcocks Road and Tantawangalo Lane at 

Tantawangalo, and at the corner of Towridgee Lane 

and South Wolumla Road. 

Roadside strips have been omitted as being too small 

to map, but can be quite significant because they 

provide an opportunity for plant species which are 

sensitive to grazing to persist. Vegetation type 21 

contains several ground cover species (for example, 

Chrysocephalum apiculatum, Calotis lappulacea) and 

one shrub, Jacksonia scoparia, which are largely 

restricted to both this vegetation type, and to 

roadsides. 

Some patches which have been mapped as remnant 

vegetation in the vicinity of Kameruka are in fact the 

large plantings of exotic trees along the Candelo 

Road. 

In general, the mapping for vegetation type 21 is not 

very useful at the fine scale of operation. Some 

patches are correctly mapped, such as along 

Bemboka River near Morans Crossing, along Sandy 

Creek between Bemboka and Candelo. Many 

patches mapped as 20 around Candelo should 

probably be 21, although some sample plots 

definitely identified as 20 come from the Candelo 

area. Some remnants which would be 21 have been 

omitted altogether. Extensive remnant patches 

mapped around the valley margins (for example, 

south and south-west of Bemboka, north-east of 

Morans Crossing, north-east of Myrtle Mountain 

and south of Kerrisons Lane near Bega) are more 

likely to be 20, or some other type. 

118 

One small patch of 21 is mapped outside the Bega 

valley, on New Buildings Road between Wyndham 

and Rocky Hall. This was checked, and although the 

main tree indicator species (E. tereticornis, 

E. melliodora, Angophora floribunda) are not present, 

snow gum (E. pauciflora) is. Snow gum is scattered 

through the drier central area of the Bega valley, 

often (though not always) coinciding with 21. As 

some of the grasses and herbs present on this site are 

typical of 21, it should probably be assumed that 

this, and any other occurrence of snow gum in the 

Wyndham area, does represent 21. Yellow box 

(E. melliodora) and snow gum occur at Rocky Hall, 

so although not mapped as 21, it probably should be 

assumed that remnant vegetation along the 

Towamba valley floor upstream of the New 

Buildings bridge could include 21. The presence on 

roadsides of the herb Chrysocephalum apiculatum 

and the grass Stipa scabra ssp. scabra supports this 

view. 

20 Bega Dry Grass Forest 

As discussed above, at least some of the areas 

mapped as 21 and 18 are probably really 20, though 

those closest to the valley margins would be more 

likely to be some other non-rain shadow valley forest 

type. 

Few extensive areas of 20 are mapped. It is mostly 

scattered in small patches, in a mosaic with 19 

throughout the Cobargo–Quaama area, Brogo, and 

the eastern half of the Bega valley. Patches are largest 

or most abundant in the following areas: 

1. around Bimbaya, between Candelo and 

Bemboka; 

2. north-west of Wolumla (Greendale Lane/ 

Wanatta Lane/Candelo–Wolumla Road); 

3. north-east of Wolumla, off Old Mill Road; 

4. adjacent to the northern edge of South East 

Forests National Park (Tantawangalo Section) 

west of Myrtle Mountain; 

5. north-east of Reedy Swamp Road, east of Bega; 

6. Goalen Head south of Murrah Lagoon; 

7. the lower Towamba valley.


Some of these areas certainly contain 20, but the 

extent has been exaggerated. For example, in the 

Greendale/Wanatta Lane remnant, 19 is also present 

(though not mapped) and a substantial amount of 

the remnant consists of silvertop–stringybark forest 

on a rocky metasediments outcrop. On the other 

hand, remnants along the Candelo–Wolumla Road 

which do represent 20 have not been mapped due to 

tree cover being too sparse. The area off Reedy 

Swamp Road is incorrectly located. Though some 20 

does occur in the vicinity, it is located in an area that 

is mapped as 60, Floodplain Wetland, while the area 

mapped as 20 is a non-rain shadow valley forest 

type. 

In the Towamba valley no evidence of 20 occurring 

was found at all. Most of the remnant vegetation in 

the Towamba valley is mapped as being 19 on lower 

lying sites or 35 (Escarpment Dry Grass Forest) on 

ridges. This appears largely correct, apart from 

possible occurrences of 21 at Rocky Hall. A species 

list was available for two sites mapped as 20, and 

suggested that the sites were either 19 or a mixture of 

19 and 35. 

Coastal occurrences of 20 have not been checked. 

While forest red gum (E. tereticornis), one of the key 

diagnostic species, does occur on the coast 

occasionally, it is often in species assemblages which 

do not suggest 20. That mapped as 20 at Goalen 

Head is on an area of better soils derived from 

gabbro, an igneous rock. There are occurrences of 

remnant forest red gum at Tanja on syenite, another 

igneous rock, and it is possible that these remnants 

do represent 20. 

Some small occurrences are mapped in reserves 

around the Bega valley margins. There is no evidence 

to support that mapped in Biamanga National Park 

in the vicinity of the Brogo Pass. The small areas 

mapped in South East Forests National Park 

(Bemboka Section) near Desert Creek do contain 

some forest red gum but the other species present 

suggest W5 (Wadbilliga Gorge Dry Forest) rather 

than 20. Similarly, while small areas within Bournda 

Nature Reserve contain forest red gum, it is unlikely 

that they represent 20. It is very likely that there is 

no 20 in reserves. 

In summary, some areas mapped as 20 contain none, 

or rather less than is mapped. This is balanced by 

some areas which are mapped as 18 or 21 and are 

more likely to be 20, particularly round the western 

edge of the Bega valley (for 21) and the Brogo area 

(for 18). There are also a large number of small 

remnants mapped as 20 which are probably correct. 

60 Floodplain Wetland 

Mapping of 60 has been based on the presence of 

riverine alluvium on the geology GIS layer, and this 

is unfortunately not very accurate. This has led to 

anomalies; for example, 60 being mapped as 

occurring on steep slopes adjacent to waterways. 

Changing the decision rules to exclude land that is 

not flat from carrying 60 will eliminate some of the 

more obvious errors in Version 4 of the map but it is 

likely that the basic inaccuracy of alluvium mapping 

will remain a problem. Discrimination between 

riverine alluvium and coastal sands also appears to be 

poor, resulting in vegetation types 61 (Coastal 

Scrub) and 36 (Dune Forest) often being 

confounded with 60 (Floodplain Wetland) and 63 

(Estuarine Wetland Scrub) around coastal lakes and 

estuaries. Another problem with mapping of 60 is 

that it sometimes does not discriminate areas of 

open water or reeds from cleared lands. This results 

in some wetlands appearing as cleared land, fringed 

by 60 on the surrounding slopes (for example, 

Whelan Swamp near Kiah). 

In order to check how well mapped instances of 60 

correlated with known wetlands, Council’s GIS layer 

containing wetlands listed in State Environmental 

Planning Policy 14 was overlaid on the CRA 

vegetation map. In many cases overlap was good, 

although generally with vegetation types 63 

(Estuarine Wetland Scrub) or 64 (Salt Marsh) rather 

than 60. This is to be expected since many SEPP 14 

wetlands are located on coastal saline lagoons. In a 

few cases SEPP 14 wetlands are mapped as either 

cleared or as various combinations of eucalypt forest 

types, or occasionally as 61 or 36, for reasons 

outlined above. 

Because of the linking of 60 to areas of alluvium, 

some quite large areas have been mapped as 60. 

Examples are south of Middle Lagoon in Mimosa 

Rocks National Park and between Kalaru and the 

north-western end of Wallagoot Lake. Checking in 

these areas revealed that there are patches of 60 along 

drainage lines or on low-lying flats, but that much of 

119


the vegetation of these areas consists of dry eucalypt 

forest types, often 36 (Dune Forest) or 46B 

(Lowland Dry Shrub Forest), both of which 

typically, or frequently, occur on sands. Most of the 

large patches which are mapped contain 60, but do 

not consist entirely of 60. This error would have 

produced a considerable overestimate of the area of 

60 present in the Shire. 

On the other hand, areas which do occur have been 

left off the map, because they occur on alluvial 

deposits which are too small to have been mapped. 

These are the upper ends of drainage lines within the 

rain shadow valleys. Since agricultural activities have 

caused extensive erosion in drainage lines in these 

valleys, many of these small alluvial deposits have 

been flushed downstream to become sand deposits 

in the beds of the major streams. However, in some 

of those which are still intact a variety of wetland 

species assemblages persist, though others have been 

converted to pasture, or to weeds. Some can be 

detected from aerial photos, since one of the main 

species, the common reed (Phragmites australis), is 

distinctively yellow against surrounding green 

pasture. Stands of Melaleuca ericifolia are less easy to 

detect against a background of pasture. The latter are 

mostly confined to the Wolumla catchment in the 

agricultural areas, while stands of Phragmites and 

other reed or sedge associations are scattered 

throughout the rain shadow valleys. Some of these 

occurrences have been included in Version 4 of the 

map, though they are not on the earlier version held 

by Council. 

Known locations include the Upper Frogs Hollow 

Creek swamp on the outskirts of Wolumla on the 

Candelo Road, Ryans Swamp near the western end 

of Towridgee Lane, a drainage line on the property 

‘Jelgowry’ at Jellat Jellat and numerous smaller areas. 

Swampy vegetation is also re-establishing in parts of 

some streambeds in the district, but given the degree 

of disturbance on these sites, it is often dominated 

by weeds. 

36 Dune Forest 

This vegetation type is mapped as occurring on 

Holocene sand deposits within a short distance of 

the ocean. Because such sand deposits are relatively 

uncommon in the Shire, there are few large patches 

of 36 mapped. Dune forest shares this habitat with 

120 

61, Coastal Scrub, which generally occurs on the 

seaward side of the Dune Forest where exposure to 

salt-laden winds is higher. The main areas mapped as 

carrying 36 are: 

1. in North Bermagui opposite the primary school; 

2. between the mouths of Baragoot and Cuttagee 

Lakes, mostly to the east of the main road; 

3. at Picnic Point, at the mouth of Wapengo Lake 

(Mimosa Rocks National Park); 

4. from Gillards Beach south to the mouth of 

Nelson Lagoon (Mimosa Rocks National Park); 

5. between Nelson Lagoon and Mogareeka 

(partially in Mimosa Rocks National Park); 

6. scattered along the coast between Tathra and the 

mouth of Wallagoot Lake (Bournda National 

Park); 

7. from the mouth of Wallagoot Lake south to 

Bournda Lagoon (Bournda National Park); 

8. around Back Lake, Merimbula; and 

9. west of Merimbula Beach, north from about the 

airport entrance. 

Most of these areas were checked, either physically 

or against species lists held for them. Most of the 

stands mapped in National Parks were found to 

either not exist or to be rather smaller than mapped, 

which would decrease the figure of 23 per cent in 

reserves substantially (although some very small 

areas which had not been mapped were also found). 

In Mimosa Rocks National Park the extent of 36 

between Gillards Beach and the mouth of Nelson 

Lagoon has been considerably overestimated. There 

is a small patch between the northern end of Nelson 

Lagoon and the beach, but apart from this all the 

area mapped as 36 is on Ordovician metasediments 

rather than sand and carries other eucalypt forest 

types. To the north of Gillards Beach there is an area 

of regenerating 36 between the Gillards Beach 

camping area and Middle Lagoon, which is part of 

the area which has been mapped as 60 (Floodplain 

Wetland). The area mapped as 36 south of Picnic 

Point was not checked, but another small patch of 

36 which has not been mapped is present on the 

road in to Picnic Point. A small area of 36 is also 

present in the Aragunnu camping area in Mimosa


Rocks National Park, which has not been mapped. 

In Bournda National Park the large strip of 36 south 

of Wallagoot Lake was checked and found to all be 

61 (Coastal Scrub). The areas along the Kangarutha 

Track between Wallagoot Lake mouth and Tathra 

were not checked, but seem unlikely to carry 36 

given that this stretch of coastline is elevated and 

rocky without any significant sand deposits. 

The 36 mapped around Back Lake, Merimbula, 

appears not to exist. That on the southern side of the 

Merimbula Lake mouth appears to have been largely 

cleared for residential development. On crown land 

south-east of the airport, the vegetation contains 

elements of 46B as well as 36, though neither of 

these types are mapped on this site. 

The main area of 36 is on crown land between 

Baragoot and Cuttagee Lakes, and is larger than 

mapping indicates, extending right to the outskirts 

of Bermagui. There is also a relatively large area 

between the Nullica River mouth and Boydtown, 

which has not been mapped as 36. Although the 

main tree species in the Boydtown stand is 

Eucalyptus viminalis rather than bangalay, 

E. botryoides, the understorey appears typical of 36. 

A similar but smaller stand featuring E. viminalis as 

the dominant tree species occurs on private property 

on the western shore of Wallagoot Lake, on an area 

mapped as containing 60 (Floodplain Wetland). The 

Boydtown area is also mostly mapped as 60, with 

very small amounts of 36 and 61. 

Mapping has exaggerated the extent of 36, and of its 

occurrence in reserves, while failing to predict the 

occurrence of some stands. Inaccuracies in geological 

mapping are probably partly responsible for this. 

63 to 70, estuarine vegetation types 

Seagrass beds have not been included in the CRA 

vegetation map. Grey Mangrove (66), Salt Marsh 

(64) and estuarine Melaleuca stands (63) have been 

mapped from aerial photo interpretation. Salt marsh 

mapping is probably reasonably accurate, though 

some areas of salt marsh appear to have been 

wrongly interpreted as 63 or 60. Mangroves seem to 

have been substantially under-mapped. They are 

mapped as being in only three locations: in the 

Bermagui estuary, at the mouth of Boggy Creek in 

Merimbula Lake, and along the Pambula River in 

the vicinity of the Pambula racecourse. Grey 

mangrove (Avicennia marina) is actually very much 

more common than this, with stands occurring 

somewhere on nearly all coastal lakes and estuaries. 

They have possibly been omitted because the 

mapping unit is a 25-metre pixel, and most 

mangrove belts around the shores of lakes are 

narrower than 25 metres. Another possible 

contributing factor is that in southern coastal regions 

mangroves tend to be short in stature, and so do not 

stand out well from adjacent salt marsh in aerial 

photos. River Mangrove (65) was not mapped, since 

occurrences of this species are very small. 

32, 34, 46B and 49, common coastal 

forest types 

The opportunity to check the accuracy of the CRA 

vegetation map for these vegetation types in selected 

coastal areas arose during assessment of some Rural 

1(c) zones for Council. Areas assessed were located 

around Bermagui, at Bald Hills west of Pambula, at 

Yowaka and Kiah. Types 32, 34, 46B and 49 were 

the major types predicted to occur in these locations, 

and were generally found to be there. However, the 

proportions of each type present often varied 

considerably from those predicted, as did the 

distribution of the types within the sites. It was also 

generally found that small areas of different 

vegetation types which were predicted to occur in 

gullies did not in fact occur, since the gullies were 

too small to support vegetation different from that 

on the surrounding slopes. Conversely there were 

sometimes small areas of other vegetation types 

which had not been predicted present. This exercise 

generally supported the findings from other areas 

that the map is reasonably accurate at the local scale, 

but not in fine detail. 

121


Appendix 3 

Vegetation types which can be 

split into further sub-types 

Vegetation types which appear to include more than 

one species assemblage include: 

60 Floodplain Wetland 

This very heterogeneous grouping has been poorly 

sampled and consists of a number of different species 

assemblages: 

• Melaleuca ericifolia closed scrub, found on small 

coastal drainage lines and in the Wolumla 

catchment, and on Murrah River backswamps; 

• emergent eucalypts (E. longifolia or E. ovata) 

over M. ericifolia, Melaleuca squarrosa and/or 

Callistemon citrinus, mixed or in single species 

stands found on small coastal drainage lines – 

this assemblage appears to have been mapped as 

57, Lowland Swamp, in northern parts of the 

Shire, although the species composition in 

northern areas is very different from that listed 

for type 57 as it occurs in the south; 

• Phragmites australis reedbeds (for example, 

Whelan Swamp near Kiah, small Bega valley 

drainage lines); 

• various sedges and rushes (Carex appressa, 

Bolboschoenus spp., Schoenoplectus validus, 

Lepidosperma longitudinale, Juncus spp.) with a 

range of herbs and grasses tolerant of wet 

conditions, found in small drainage lines within 

agricultural areas (where these have not been 

destroyed by gully erosion), along swampy 

sections of sediment choked rivers, and a dune 

swamp west of Pambula Beach; 

• permanent and semi-permanent freshwater 

lagoons with submerged and emergent aquatic 

plants (including Eleocharis sphacelata, Typha 

orientalis, Villarsia reniformis, Potamogeton spp) 

found in backswamps of the Bega and Murrah 

floodplains, Long Swamp and Bermagui Swamp, 

Penuca Swamp and possibly in some farm dams 

122 

which are not used for irrigation and hence not 

subject to changes in water level. 

All are depleted in the area, but particularly so 

within agricultural areas, and all are of high 

conservation significance. 

19 Bega Wet Shrub Forest 

This vegetation type has three possible sub-types 

which appear to reflect differences in site quality: 

• Eucalyptus elata with the shrubs Cassinia trinerva, 

Hymenanthera dentata or Kunzea ericoides and 

sometimes Leptospermum brevipes occurs on 

steep south-facing slopes on granitoids or 

metasediments. The shrub layer is dense and 

ground cover often minimal. This species 

assemblage is found on the poorer sites, often 

with shallow rocky soils. A typical location is on 

steep slopes above Myrtle Creek near Burragate. 

• E. baueriana with E. elata and a diverse grass or 

herb ground cover in which Bergalia tussock, 

Carex longebrachiata is prominent and shrubs of 

Hymenanthera dentata and Cassinia trinerva are 

more scattered occurs on sites of high soil 

fertility and moisture reliability, such as river flats 

or gullies on granitoids or alluvium. 

• E. elata and E. baueriana may be joined or even 

replaced on some sites by E. viminalis, 

E. angophoroides and E. radiata with a mixed 

shrub and grass or herb understorey. This usually 

occurs on lower slopes not immediately 

associated with streams. This type is more 

common in valley margin sites such as in 

Murrabrine National Park and east of 

Wyndham. 

The three variants can occur in a mosaic where 

suitable conditions occur in close juxtaposition, for 

example, between Burragate Road and Myrtle 

Creek. The sub-type containing E. baueriana and 

often a less shrubby understorey is probably more 

restricted to private land tenure and more affected by 

clearing than the other two sub-types.


32 Coastal Foothills Dry Shrub 

Forest 

In this vegetation type, sites dominated by spotted 

gum (Corymbia maculata) and ironbarks 

(E. paniculata or E. tricarpa) could be separated 

from those dominated by woollybutt (E. longifolia), 

coast grey box (E. bosistoana) and ironbark 

(E. tricarpa). To the north of Bega Valley Shire the 

spotted gum–ironbark assemblage, usually with 

burrawang (Macrozamia communis) in the 

understorey, is more extensive and tends to be more 

obviously distinct from woollybutt–box–ironbark 

stands. Similarly, Coastal Gully Shrub Forest, 34, 

varies considerably, with spotted gum, woollybutt 

and ironbark dominating coastal stands north of 

Bega. Non-coastal stands on metasediments at 

Brogo, and on the coast south of Bega have a less 

diverse, less grassy understorey and tend to be 

dominated by monkey gum (E. cypellocarpa) and 

stringybarks (D Keith, pers. comm.). 

38 and 39, Riparian Scrubs 

Rather than having two Riparian Scrub types, three 

might describe the range of species combinations 

better. All three contain the core species Acacia 

floribunda, Callistemon subulatus, Leptospermum 

emarginatum and L. obovatum. 

• Southern streams (for example, Imlay Creek, 

Wallagaraugh, Yowaka and rocky sections of the 

Towamba River) also carry the distinctive group 

Melaleuca armillaris, Calytrix tetragona, Grevillea 

linearifolia and Leptospermum scoparium, 

sometimes with Tristaniopsis laurina or 

Leptospermum lanigerum. 

• Rivers with a substantial catchment above the 

escarpment (Tantawangalo Creek, Bemboka 

River) and which flow through dry rain shadow 

valleys tend to include Leptospermum lanigerum, 

Hakea microcarpa and Acacia elongata but not 

Tristaniopsis laurina. 

• Northern streams (Brogo and tributaries, 

Murrah River) have the core species plus 

Tristaniopsis laurina but none of the other 

additional species present in southern or central 

streams, possibly because in these streams the 

riparian zone is shared with Riverine Forest (40) 

dominated by Casuarina cunninghamiana. This 

tends to make these streams shadier places than 

those where casuarina is absent, and riparian 

shrub associations appear to be most diverse in 

river reaches where the bed is exposed to full sun. 

Upper reaches under forest tend to carry a much 

reduced riparian flora consisting only of the few 

shade-tolerant species. 

The sub-type characteristic of dry rain shadow 

valleys, with most of its occurrence on the Bega 

River system (excluding the Brogo) is the most 

depleted. 

46B Lowland Dry Shrub Forest 

This vegetation type contains a sub-type which 

occurs primarily on sand and includes blackbutt 

(E. pilularis) as a co-dominant tree with bloodwood 

(Corymbia gummifera). Other distinctive species are 

Banksia serrata, the tall herb Trachymene anisocarpa 

and the grasses Stipa pubescens and Anisopogon 

avenaceus. Blackbutt occurred in only two of 19 

plots referrable to 46B, so this sub-type is obviously 

considerably less common than the more typical 

46B. The main occurrence is in the Pambula Beach 

area and the northern end of Ben Boyd National 

Park. Blackbutt is at the southern limit of its 

distribution in the Shire and stands containing this 

species could be regarded as being of greater 

conservation significance than typical 46B. 

52 Mountain Rock Scrub 

This vegetation type has been described from 

occurrences on large granitoid outcrops at the 

southern end of the Shire. There are extensive areas 

of similar outcrops in Wadbilliga National Park and 

on Little Mumbulla Mountain in the north of the 

Shire. Species assemblages here have not been 

sampled, and that occurring on the summit of 

Dr George Mountain gives some indication that 

they may be different from those in the south. There 

are also rock scrubs on cliffs on Devonian 

metasandstones and conglomerate in South East 

Forests National Park (Bemboka Section) which are 

possibly different again from those on northern 

granitoids, or southern metasediments. 

123


62 Coastal Scrub 

Coastal Scrub falls fairly readily into two sub-types, 

one dominated by Banksia integrifolia on sand dunes 

and headlands formed from Tertiary sediments, and 

another dominated by Allocasuarina verticillata or 

Melaleuca armillaris on rocky headlands. Almost 

mono-specific stands of Melaleuca armillaris are 

sometimes found in other near-coastal situations 

besides headlands, and may be a result of wildfire. 

6 Coastal Warm Temperate 


This vegetation type includes another rainforest 

assemblage occurring in Bega Valley Shire which has 

not been sampled or described. It is dominated by 

124 

the tree Backhousia myrtifolia, with a very sparse 

understorey consisting of ferns and the large tussock 

plant Lepidosperma urophorum. Epiphytes such as 

bird’s nest fern (Asplenium australasicum) and various 

orchids (Sarcochilus spp, Plectorrhiza tridentata) 

occur on trees or rocks. This vegetation type is found 

in less sheltered gullies than warm temperate 

rainforest, generally at low altitudes north of Bega. It 

is much more common beyond the Shire to the 

north. All known occurrences are on Ordovician 

metasediments. Examples occur on the lower slopes 

of Peak Alone below Paddy’s Creek Flora Preserve, in 

Bermagui and Murrah State Forests, Biamanga and 

Wadbilliga National Parks. Individual stands tend to 

be quite small.


Appendix 4 

Environmental weeds of Bega 


The following table includes species which are 

common and widespread weeds in the Shire, weeds 

with only localised distributions and species which 

are serious environmental weeds in other districts of 

New South Wales or Victoria but, as yet, not well 

established in Bega Valley Shire. The latter may have 

the potential to become more abundant here in 

time. 

An environmental weed could be defined as any 

plant not naturally occurring in an area which 

invades native vegetation. This very wide definition 

means that many agricultural weeds, and even useful 

pasture species, may also be regarded as 

environmental weeds where they are growing in 

remnant native vegetation rather than in improved 

pasture or crops. 

Table 6: Environmental weeds of Bega Valley Shire 

Scientific name Common name Widespread Localised Not yet 

established 

Trees 

Acacia baileyana Cootamundra 

wattle 

It is not within the scope of this report to make 

recommendations about the management of 

environmental weeds within the district. With 

increasing awareness of the effects of environmental 

weeds there may be moves at State or National level 

to declare some horticultural species as noxious, as 

occurred recently for most willow varieties. There is 

also increasing awareness in the horticulture industry 

of the weed potential of some ornamental plants, 

with the result that some nurseries no longer sell 

some of these plants. The question of the climatic 

conditions required for certain species to become 

weedy needs resolving, since some species may be 

extremely invasive in some regions and quite benign 

in others. 

In Table 6, species which are declared noxious in the 

Shire have (N) listed after the common name. 

Comments 

X Becoming established on 

roadsides, occasionally in bush if 

planted near it. Likely to greatly 

increase after fire, as may other 

non-local wattles being planted on 

rural properties (for example, 

A. podalyriifolia, A. saligna). 

Acer negundo box elder X Regenerates freely along stream 

banks, for example, Bega River 

near Tarraganda bridge, and on 

nearby lagoons. 

Ailanthus altissima tree of heaven X Suckering habit allows this plant 

to form dense thickets. Capable of 

invading otherwise undisturbed 

bush. 

Alnus spp. alders X Not yet established in the district, 

but increasingly planted in 

gardens and has shown the ability 

to regenerate freely from seed in 

riparian situations in Canberra. 

125


Table 6: Environmental weeds of Bega Valley Shire (continued) 


established 

Chamaecytisus 

palmensis 

Crataegus 

monogyna 

126 

tagasaste or 

tree lucerne 

Ligustrum lucidum large-leafed 

privet 

Pinus radiata radiata or 

Monterey pine 

X Has been widely planted on a 

small scale for stock fodder and 

amenity plantings. Reproduces 

freely from seed, but palatability 

of seedlings to livestock, rabbits 

and native herbivores ensures that 

few survive. Where grazing is 

excluded (for example, within 

towns) it has successfully 

naturalised. Invasive in remnant 

vegetation in some parts of 

Victoria and New South Wales. 

hawthorn X Mostly seen in paddocks in drier 

areas, for example, Candelo. 

X Common and widespread, near 

creeks or in gullies. 

X Readily invades bush from 

plantations or windbreaks nearby. 

Big potential for this in Pericoe 

area, and frequency may increase 

elsewhere as more pine 

plantations are established in the 

district. 

Populus alba white poplar X Suckering habit creates large 

thickets. Occasionally seen in 

moist sites such as gullies and river 

banks. Has been planted for 

erosion control to a small extent. 

Populus nigra var 

italica 

Robinia 

pseudoacacia 

Lombardy 

poplar 

X Mostly in river beds, where it 

suckers and forms thickets which 

may obstruct flows. 

black locust X Suckering habit creates large 

thickets. 

Salix spp. willows (N) X X Only a problem in rivers where 

dense growth in the beds can 

obstruct flows leading to bank 

erosion. Recently declared a 

weed of national significance 

(except for weeping willows and 

some pussy willows). 

Schinus areira 

(syn. S. molle) 

Comments 

pepper tree X Occasionally naturalised in the 

Candelo area, rarely within 

remnant vegetation. Invasive in 

remnant vegetation on the New 

South Wales western slopes. 

Ulmus procera English elm X Suckering habit can create large 

thickets.




established 

Climbers 

Anredera 

cordifolia 

Araujia sericiflora 

(syn. A hortorum) 

Delairea odorata 

(syn. Senecio 

mikanioides) 

Madeira vine X Not yet well established, but 

present on river banks in Candelo 

and Bega. 

moth plant X Sparsely distributed on riverbanks 

and in wetter bush. 

Cape ivy X Widely distributed smothering 

vine. 

Dipogon lignosus dolichos pea X Sparsely distributed smothering 

vine. 

Hedera helix English ivy X Generally only in towns or close 

to buildings. Can smother trees. 

Lonicera japonica Japanese 

honeysuckle 

Asparagus 

asparagoides 

(syn. Myrsiphyllum 

asparagoides) 

Protasparagus 

spp. 

Shrubs 

Chrysanthemoides 

monilifera 

bridal veil 

creeper, florist’s 

smilax 

X The most common climbing 

environmental weed in the district, 

usually in shady sites or close to 

water. 

X Very invasive small creeper which 

tends to behave more as a ground 

cover than a climber. Can blanket 

the ground and exclude most 

other plants. Mostly coastal but 

also in drier woodland remnants. 

Recently declared a weed of 

national significance. 

asparagus fern X Not common except in towns, 

particularly Merimbula. 

bitou bush (N) X Largely restricted to the Tathra 

area, where control efforts have 

made a substantial impact. 

Vigilance is needed in other 

coastal areas as small outbreaks 

have been found at Cuttagee, 

Pambula Beach and Eden. 



Coprosma repens mirror bush X Not common, usually close to 

beaches. Has shown the capacity 

to behave more invasively in 

Eurobodalla Shire immediately to 

the north and in Merimbula. 

Cotoneaster 

glaucophyllus 

Cytisus scoparius English broom 

(N) 

Comments 

cotoneaster X Widespread, on roadsides, 

riverbanks, waste ground in 

towns. 

X On roadsides in Candelo village. 

Very invasive in moist forest at 

highter elevations in New South 

Wales and Victoria. 

127




established 

Genista linifolia flaxleaf broom X On roadsides in Candelo village. 

Genista 

monspessulana 

128 

Montpelier 

broom (N) 

X One infestation within Eden town 

boundaries, which repeated 

spraying by Council has not yet 

eliminated. Also scattered on 

urban fringes and roadsides 

throughout the Shire. Shows a 

high potential to become more 

widespread. 

Lantana camara lantana X Sparsely distributed at Wallaga 

Lake and as far south as Tathra. 

Common north from Tilba Tilba, 

but drier climate and poorer soils 

may prevent it becoming common 

further south. Recently declared a 

weed of national significance. 

Leptospermum 

laevigatum 

Ligustrum sinense small-leaved 

privet 

Lycium 

ferocissimum 

Opuntia 

aurantiaca 

Opuntia stricta var 

stricta 

coastal tea tree X This native species occurs both 

north and south of the Shire, but 

appears to be naturally occurring 

within the Shire only in the far 

south (for example, Green Cape). 

It has been widely planted in 

coastal areas where it has begun 

spreading into coastal banksia 

scrub, for example, Tathra. 

African 

boxthorn 

X Common and widespread, 

usually in gullies or creek-side 

sites. 

X Generally under trees in paddocks 

in drier parts of the district such as 

Candelo, where it may degrade 

remnant grassy forest. 

tiger pear (N) X One infestation known from dry 

rocky site in Brogo. 

common prickly 

pear (N) 

X Not common; dry rocky sites at 

Burragate and Rocky Hall and on 

riverbank, lower Bega River. 

Polygala myrtifolia milkwort X Invades coastal bush around 

towns, for example Pambula 

Beach, Merimbula. 

Psoralea pinnata African scurf 

pea 

Pyracantha 

angustifolia 

pyracantha, 

firethorn 

Rosa rubiginosa sweet briar, 

briar rose 

Comments 

X Invades coastal bush around 

towns, for example Bermagui, 

Pambula racecourse. 

X Occasionally found on urban 

fringes, but as the seed is 

bird-distributed it has the potential 

to become more widespread. 

X Mostly in cleared paddocks, can 

occur in remnant grassy forest, but 

not very invasive in this district.




established 

Rubus ulmifolius Blackberry (N) X One of the worst environmental 

weeds in the district, though it 

mostly invades disturbed rather 

than intact bush. Can choke out 

all other plants over large areas. 



Solanum 

mauritianum 

wild tobacco 

bush 

X Very invasive in shady moist forest 

in Eurobodalla and further north. 

Ulex europaeus gorse X Not common; a few plants 

present at Kameruka and 

Bemboka. A very invasive weed 

in cool tableland climates. 

Declared a weed of national 

significance. 

Water plants 

Eichhornia 

crassipes 

Isolepis prolifera no common 

name 

water hyacinth X A single infestation occurred (and 

was eliminated by repeated 

spraying over several years) on 

the Glebe Lagoon in Bega. While 

present it spread rapidly, showing 

a high potential to degrade water 

bodies in the Shire. 

X Common in disturbed wetlands 

where it displaces natives. 

Juncus articulatus jointed rush X Very common in local streams 

where it forms dense mats which 

may exclude natives. Probably 

helps protect against erosion 

however. 

Sagittaria 

graminea 

sagittaria X Recently found growing on Brogo 

Dam margins, where chemical 

control has been implemented to 

prevent its spread. Weed of 

irrigation areas. 

Salvinia molesta salvinia (N) X Currently only known from a 

single farm dam at Bunga, where 

it has proliferated rapidly. Shows 

high potential to degrade water 

bodies in the Shire. Troublesome 

in Eurobodalla Shire. 

Herbs 

Ageratina 

adenophora 

Crofton weed 

(N) 

Comments 

X A highly invasive weed of forest 

margins and clearings normally 

found north of about the 

Shoalhaven district. Infestations 

have occurred around Eden, which 

Council and NPWS have been 

attempting to eliminate. A small 

outbreak has also appeared on the 

road verge on Brown Mountain. 

129




established 

Artemisia 

verlotiorum 

130 

Chinese 

mugwort 

(‘ragweed’ 

locally) (N) 

X In lower Bega River where it forms 

dense stands on banks and 

islands, excluding all other 

species. 

Cakile maritima sea rocket X Common on beaches, where it 

assists with native species in 

stabilising the fore-dune. 

Cirsium vulgare black thistle X Can spread into remnant grassy 

open forest from pasture. 

Conium 

maculatum 

hemlock X Can spread into forest in shady or 

riparian sites. 

Conyza albida fleabane X Spreads freely into disturbed bush 

from pasture. 

Daucus carota wild carrot X Currently found mostly on 

highway verges throughout the 

Shire, but has proved capable of 

invading less disturbed 

vegetation. 

Euphorbia lathyrus caper spurge X Generally on riverbanks where it 

may form dense stands. 

Euphorbia paralias sea spurge X Found on beaches south of 

Narooma (only New South 

Wales records so far). A 

troublesome weed in other States. 

Hydrocotyle 

bonariensis 

Hypericum 

perforatum 

Hypochaeris 

radicata 

Leontodon 

taraxacoides 

Plantago 

lanceolata 

pennywort X Common on beaches, where it 

assists with dune stabilisation. 

St John’s wort 

(N) 

cat’s ear or 

flatweed 

Comments 

X Has been patchily present on 

roadsides in the district for some 

years, but spread into adjacent 

agricultural areas has been 

slowed by Council’s control 

efforts. Invades remnant bush and 

native grassland on the Monaro. 

X Ubiquitous weed absent only from 

the most pristine bush, but seldom 

becomes abundant enough in 

bush to exclude native plants. 

Does so in disturbed grassy open 

forest. 

lesser hawkbit X A very similar plant to cat’s ear, 

but preferring wetter sites. May 

become abundant on swamp 

margins, as well as invading 

remnant grassy vegetation on 

roadsides and in paddocks. 

plantain X Commonly invades native pasture 

and disturbed grassy open forest.




established 

Ranunculus repens creeping 

buttercup 

Rorippa nasturtiumaquaticum 

Tradescantia 

albiflora 

X Occurs on creek banks and in 

gullies, where it may occasionally 

become dense enough to 

displace native wetland 

vegetation. 

watercress X Found in gullies and creeks where 

it can form dense mats. 

wandering jew X In shady sites close to water, 

where it can form dense mats, 

smothering native ground covers 

and preventing tree and shrub 

regeneration. 

Trifolium spp. clovers X Can invade remnant grassy open 

forest from nearby pasture. 

Verbena rigida verbena X Can invade remnant grassy open 


Vinca major periwinkle X Spreads along streams (even into 

undisturbed sites) and forms a 

dense ground cover, excluding all 

other plants. A serious riparian 

weed which has the potential to 

prevent regeneration of native 

riparian shrubs. However, 

probably helps prevent erosion in 

the short term. 

Zantedeschia 

aethiopica 

Grasses 

Ammophila 

arenaria 

Andropogon 

virginicus 

arum lily X More common north from Tilba 

Tilba but occasionally established 

near habitation in the Shire and 

has the potential to smother native 

vegetation along drainage lines. 

marram grass X Has been planted to stabilise 

dunes, but native species would 

be preferable. 

whisky grass X Currently found mostly on road 

verges throughout the Shire, but 

has proved capable of invading 

less disturbed vegetation. 

Dactylis glomerata Cocksfoot X A pasture grass which often 

invades moist or nutrient-enriched 

native pasture, particularly on 

roadsides. 

Ehrharta erecta panic 

veldtgrass 

Comments 

X Dense grass which invades 

disturbed coastal forests, for 

example, Merimbula vacant 

crown land. 

131




established 

Eragrostis curvula African 

lovegrass (N) 

132 

X Principally a weed of pasture, but 

spreading along roadsides and 

well established in bush in several 

parts of the district. Serious weed 

of remnant grassy open forest and 

has shown the ability to invade 

bush (mostly along track verges) in 

dry sandy soils on the coast, for 

example, Tura Beach. 

Holcus lanatus Yorkshire fog X Forms dense stands in damp or 

shady grassland sites, choking out 

native grasses and herbs. 

Nassella 

trichotoma 

Paspalum 

dilatatum 

serrated tussock 

(N) 

X Currently mostly in Towamba 

valley, but with smaller infestations 

elsewhere and potential to spread 

within agricultural areas over the 

whole district. Will grow under 

light forest and has the potential to 

degrade grassy open forest 

remnants. Recently declared a 

weed of national significance. 

paspalum X Forms dense stands in damp, 

sunny sites, degrading wetlands. 

Phalaris aquatica phalaris X A pasture grass which often 

invades native pasture and 

roadside remnant vegetation. 

Pennisetum 

clandestinum 

Phyllostachys 

aurea 

Sporobolus indicus 

var capensis 

Stenotaphrum 

secundatum 

kikuyu X Spreads gradually by vegetative 

means into remnant grassy open 

forest and wetlands, where it can 

suppress most other species. Does 

not like shade so less troublesome 

in bush. 

golden 

bamboo 

Parramatta 

grass 

Comments 

X Large infestation on Bega river 

bank, where it excludes all other 

species. Reproduction from seed 

unlikely. 

X Invades remnant grassy open 


buffalo grass X Mostly on coastal headlands 

where it forms a dense ground 

cover excluding all other species. 

Not common. 

Vulpia bromoides rat’s tail fescue X A pasture grass which often 

invades native pasture and 

roadside remnant vegetation.


Appendix 5 

Threatened and rare plants of Bega Valley Shire 

Conservation status: 

V, E = vulnerable, endangered in Commonwealth Endangered Species Protection Act 1992. 

v, e = vulnerable, endangered in Threatened Species Conservation Act 1995 (New South Wales). 

3RCa = ROTAP* code (2 = distributional range 100 km 

R = rare but not threatened in Australia 

C = represented in a conservation reserve 

a = >1000 plants in a reserve 

i =


Table 7: Threatened and rare plants of Bega Valley Shire (continued) 

Species Conservation status Approximate location recorded 

Eucalyptus imlayensis E, e Mount Imlay National Park 

Eucalyptus spectatrix 2RC– Dr George Mountain, SEFNP Bemboka Section 

Genoplesium rhyoliticum E, e Nullica rhyolite 

Grevillea acanthifolia ssp. 

paludosa 

134 

E, e Nalbaugh Plateau (SEFNP Genoa Section) 

Grevillea irrasa ssp. irrasa ms may be nominated Nullica rhyolite area (newly described member of 

Grevillea victoriae complex) 

Haloragodendron bauerlenii 3RCa Brogo Pass, Dr George Mountain 

Hibbertia hermannifolia 3RCa Mount Poole, SEFNP Tantawangalo Section 

Lepidium hyssopifolium E, e Wallaga Lake National Park 

Myoporum bateae 3RC– Mount Pericoe, Murrah, Mimosa Rocks National 

Park 

Persoonia brevifolia 2RCa Mount Imlay National Park, SEFNP Genoa Section 

Phebalium carruthersii 3RC– Dr George Mountain, Mumbulla Mountain, Mount 

Pericoe 

Phebalium ralstonii V, v Nullica rhyolite 

Phebalium rhytidophyllum V, v Nalbaugh Plateau (SEFNP Genoa Section) 

Pomaderris brogoensis 3RC– Brogo Pass, SEFNP Bemboka Section, Wadbilliga 

NP, Nadgee NR, occasional on private property 

around valley margins 

Pomaderris costata 3RC– Jingera Rock, Letts Mountain, Mount Poole, Brown 

Mountain, Pheasants Peak 

Pomaderris cotoneaster E, e Tantawangalo Creek above weir, Coolangubra 

Pomaderris elachophylla e Brown Mountain 

Pomaderris parrisiae V, v Brown Mountain 

Pomaderris pauciflora 3RC– Nungatta, New Buildings Road 

Pomaderris virgata 3RC– Dr George Mountain, Burragate Peak, Mount Poole 

Prasophyllum affine E, e Nadgee Nature Reserve 

Prostanthera walteri 3RCa Mount Imlay National Park, Nalbaugh Plateau 

Pseudanthus divaricatissimus 3RCa Nullica rhyolite and Mount Poole 

Pseudanthus ovalifolius e Ben Boyd National Park 

Pultenaea pedunculata e Bournda National Park and Cuttagee 

Rulingia hermannifolia 3RCa Nullica rhyolite, Nethercote Falls Flora Reserve 

Senecio spathulatus e Nadgee Nature Reserve? 

Senecio squarrosus e Nadgee Nature Reserve? 

Spyrideum cinereum 3RCa Nadgee Nature Reserve


Table 7: Threatened and rare plants of Bega Valley Shire (continued) 

Species Conservation status Approximate location recorded 

Thesium australe V, v Snowy Mountains Highway verge, Numbugga, 

in kangaroo grass 

Viola cleistogamoides e Wonboyn (single old record) 

Westringia davidii V, v Nullica rhyolite 

Wilsonia backhousei v Salt marsh, Bermagui and Wapengo 

Wilsonia rotundifolia e Salt marsh, present in East Gippsland and on New 

South Wales coast but not yet recorded in Bega 


Zieria sp. G (buxijugum ms) E, e Lochiel, private property 

Zieria sp. H (formosa ms) E, e Lochiel, private property 

Zieria sp. I (parrisiae ms) E, e Lochiel, private property 

Note: SEFNP = South East Forests National Park 

135


Appendix 6 

Species of regional conservation significance 

The following tables list species of regional conservation significance, including regionally uncommon species 

and those for which habitat is depleted, which are likely to occur on private property. 

Table 8: Species of rain shadow valley dry grass forests 

Species Habitat 

Trees 

Eucalyptus baueriana Bega Wet Shrub Forest, 19 

Eucalyptus dives usually Bega Dry Grass Forest, 20 

Eucalyptus melliodora Candelo Dry Grass Forest, 21 

Eucalyptus pauciflora usually Candelo Dry Grass Forest, 21 

Eucalyptus tereticornis vegetation types 18, 20 and 21, rarely other types on the coast 

Shrubs 

Deeringia amaranthoides Brogo Wet Vine Forest,18 

Jacksonia scoparia usually Candelo Dry Grass Forest, 21 

Grasses 

Austrostipa scabra ssp. scabra Candelo Dry Grass Forest, 21 

Austrostipa verticillata Candelo Dry Grass Forest, 21 

Capillipediun parviflorum Candelo Dry Grass Forest, 21 

Cenchrus caliculatus Brogo Wet Vine Forest, 18 

Chloris truncata vegetation types 20 or 21 

Chloris ventricosa Candelo Dry Grass Forest, 21 

Sporobolus elongatus usually Candelo Dry Grass Forest, 21 

Paspalum orbiculare ? Bega Dry Grass Forest, 20 

Graminoids 

Cyperus sphaeroideus Bega Wet Shrub Forest, 19 

Cyperus imbecillis Bega Wet Shrub Forest, 19 

Eriocaulon scariosum Towamba Cemetery 

Scleria mackaviensis vegetation types 20 or 21 

Herbs 

Laxmannia gracilis vegetation types 20 or 21 

Bulbine bulbosa vegetation types 20 or 21 

Bulbine glauca vegetation types 20 or 21 

136


Table 8: Species of rain shadow valley dry grass forests (continued) 


Wurmbea biglandulosa vegetation types 20 or 21 

Microtis unifolia vegetation types 19, 20 or 21 

Brachycome ciliaris var. ciliaris vegetation types 20 or 21 

Calotis lappulacea Candelo Dry Grass Forest, 21 

Chrysocephalum apiculatum Candelo Dry Grass Forest, 21 

Leptorhynchos squamatus ssp. A Candelo Dry Grass Forest, 21 

Leucochrysum albicans ssp. albicans Candelo Dry Grass Forest, 21 

Triptilodiscus pygmaeus vegetation types 20 or 21 

Vittadinia cuneata var. cuneata forma 

cuneata 

vegetation types 20 or 21 

Vittadinia dissecta vegetation type 18 

Vittadinia muelleri vegetation types 20 or 21 

Cullen microcephalum vegetation type 21 

Glycine tomentella sens. lat. vegetation types 20 or 21 

Lespedeza juncea var. sericea vegetation types 20 or 21 

Lotus australis Candelo Dry Grass Forest, 21 

Zornia dyctiocarpa var. dyctiocarpa vegetation types 19, 20 or 21 

Velleia paradoxa Candelo Dry Grass Forest, 21 

Linum marginale vegetation types 20 or 21 

Veronica gracilis Wyndham Cemetery 

Polygala japonica vegetation types 19, 20 or 21 

Calandrinia pickeringii shallow soil around granite outcrops, Burragate, Brogo 

Pimelea curviflora var. sericea Candelo Dry Grass Forest, 21 

Ferns 

Botrychium australe vegetation type 19 

Davallia pyxidata granite boulders, usually Brogo Wet Vine Forest, 18 

Table 9: Riparian or wetland species 


Trees 

Casuarina cunninghamiana Riverine Forest, 40, north of Bega 

Shrubs 

Acacia elongata Northern Riparian Scrub, 39, Tantawangalo Creek and Bemboka 

River 

Amyema cambagei mistletoe of Casuarina, Narira Creek catchment only? 

137


Table 9: Riparian or wetland species (continued) 


Grasses 

Amphibromus nervosus ? farm dam margins 

Isachne globosa creek beds within agricultural areas 

Panicum bisulcatum muddy margins of river bed, Brogo Pass 

Graminoids 

Carex brownii swampy drainage lines 

Cyperus sanguinolentus drainage lines and seepage areas within agricultural areas 

Cyperus gunnii swamps 

Cyperus laevigatus estuarine locations 

Cyperus lhotskyanus stream banks and swamps 

Lepidosperma longitudinale dune swamps on sand, for example, Long Swamp, Pambula 

Beach swamp 

Herbs 

Ottelia ovalifolia lagoons, occasionally farm dams 

Philydrum lanuginosum lagoons, for example, Long Swamp, Bermagui Swamp 

Sonchus hydrophilus drainage line on agricultural land, Quaama 

Callitriche muelleri wet shady places, rainforest gullies, Melaleuca swamp 

Crassula helmsii margins of water bodies, lagoons, farm dams 

Elatine gratioloides margins of water bodies, lagoons, farm dams 

Haloragis heterophylla wet grassland in drainage lines 

Villarsia exaltata lagoons, coastal Melaleuca swamps 

Villarsia reniformis lagoons, for example, Long Swamp, Penuca Swamp 

Muehlenbeckia gracillima river beds 

Persicaria praetermissa coastal Melaleuca swamp 

Lysimachia japonica Melaleuca swamp, Jellat Jellat 

Samolus valerandi creek beds in agricultural areas and valley margins 

Mazus pumilio waterlogged ground in Melaleuca swamp or Swamp Forest, 58 

Mimulus repens coastal swamp 

Calystegia sepium wetland margins 

Viola caleyana drainage lines in agricultural areas and on the coast 

138


Table 10: Species known mostly from private property around rain shadow valley margins 


Trees 

Cryptocarya microneura rainforest at Murrabrine 

Symplocos thwaitesii rainforest east of Bega 

Shrubs 

Adriana tomentosa var. tomentosa disturbed areas near rivers, Brogo and Bega 

Bertya rosmarinifolia near Tantawangalo Creek 

Omalanthus stillingifolius steep rocky sites, Brogo Pass and Brogo Dam 

Senna aciphylla steep rocky sites, Brogo Pass and Burragate 

Pomaderris bodalla steep rocky sites near Brogo Pass and riparian, Dignam’s Creek 

Pomaderris brogoensis steep rocky sites, Brogo Pass and nearby, scattered along creeks 

and on rainforest margins 

Pomaderris phylicifolia var. ericoides dry rocky ridge at Brogo 

Pomaderris pilifera Brogo River banks 

Korthalsella rubra mistletoe of rainforest trees 

Notothixos subaureus mistletoe parasitic on other mistletoes 

Grasses 

Austrostipa ramosissima margins of rainforest 

Vines 

Cissus antarctica rainforest, north of Bournda 

Sicyos australis wet gullies, often in vegetation types 18 or 19 

Herbs 

Aneilema biflorum shady wet gullies 

Lepidium pseudotasmanicum Brogo Wet Vine Forest, 18 

Chenopodium carinatum Escarpment Dry Grass Forest, 35 

Chenopodium pumilio ? edges of agricultural areas 

139


Table 11: Species occurring mostly on the coast and known only from privately owned land or vacant 

crown land or with a significant proportion of records from private property with some occurrences in 

reserves as well 


Trees 

Cryptocarya glaucescens rainforest, Murrah 

Ficus obliqua rainforest, Goalen Head and Mimosa Rocks National Park 

Shrubs 

Aegiceras corniculatum river mangrove, Bermagui River and Merimbula Lake 

Astrotricha sp. aff. ledifolia Bournda National Park, vacant crown land and private property 

along Sapphire Coast Drive 

Grevillea mucronulata Tura Beach area 

Phebalium squamulosum ssp. argenteum coastal cliffs, Bermagui and Nadgee 

Solanum stelligerum disturbed forest margins, Bermagui, Murrabrine 

Herbs 

Trachymene anisocarpa Pambula Beach crown land and northern Ben Boyd National Park 

Calocephalus lacteus around swamp on private property near Baragoot Lake, only 

New South Wales record? – northern limit 

Polymeria calycina in native pasture or grassy understorey, Bermagui, Bournda 

National Park 

Limonium australe estuarine, Merimbula Lake, Pambula racecourse wetlands 

Muehlenbeckia rhyticarya Dr George Mountain 

Stackhousia spathulata forest edges near estuaries and coastal lakes 

140


Appendix 7 

Remnant vegetation in Bega 

Valley Shire on roadsides and 

other public lands 

administered by Council 

It has been calculated that there are approximately 

3 120 000 hectares of rural roads in New South 

Wales, an area roughly equivalent to that of National 

Parks. Roadsides potentially have a lot to contribute 

to conservation of native vegetation. In some 

agricultural areas in New South Wales roadsides 

represent the main location of remnant vegetation. 

This is less the case in Bega Valley Shire, where 

substantial amounts of native vegetation remain in 

public ownership, in State Forests, National Parks 

and other reserves. However, those vegetation types 

in the Shire which occurred solely or primarily in 

areas which have been cleared for agriculture, are 

now represented only by small fragmented 

remnants. Some of these occur on roadsides and on 

other public lands such as cemeteries, showground 

reserves and travelling stock reserves, as well as on 

private property. Road verges in Bega Valley Shire 

tend not to be very wide, unlike western New South 

Wales where many double as travelling stock routes, 

and this limits their usefulness as habitat for remnant 

vegetation and fauna. 

There are several reasons why a roadside remnant 

might be significant: 

1. it may contain species which are formally 

recognised as being threatened (listed in the 

Threatened Species Conservation Act 1995 for 

New South Wales, or the Endangered Species 

Protection Act 1992 for Australia); 

2. it may contain species which, while not regarded 

as threatened on a national or State scale, are 

regionally or locally rare; 

3. it may represent a good example of a vegetation 

community which has been largely removed or 

degraded by agricultural activity; 

4. it may represent habitat for threatened fauna, or 

simply for more common fauna which could not 

continue to survive in agricultural areas without 

the presence of such remnants; 

5. it may serve to connect other remnants, thus 

increasing the possibilities of genetic exchange 

between otherwise isolated populations of plants 

and animals, and hence improving their chances 

of survival in the long term. 

This report presents some information about the 

existence of such roadside vegetation in the Bega 

Valley Shire based on casual observations over several 

years, and on a brief survey of some roadsides and 

cemeteries conducted in October 1999. It should 

not be considered to be complete, although the brief 

survey did concentrate on the areas considered most 

likely to contain significant species or vegetation 

types. A more thorough survey is still desirable, and 

should preferably be combined with a survey of 

other potentially significant public lands such as 

travelling stock reserves, showground reserves, 

cemeteries and vacant crown land. It is essential that 

such a survey be done in spring or early summer, 

since many herbs and grasses which are characteristic 

of the largely cleared grassy forest ecosystems of the 

agricultural areas can only be detected when 

flowering or seeding at these times. 

1. Threatened plant species 

No species listed as threatened in the Threatened 

Species Conservation Act 1995 have been found to 

occur on roadsides in the district. 

There are two ROTAP (Briggs & Leigh 1995) 

species which do occur on roadsides, although both 

occur in areas where native vegetation is not 

confined to the roadside. They are: 

• Eucalyptus badjensis, a large tree which has a 

restricted distribution along the coastal 

escarpment, including the area between Brown 

Mountain and Mount Darragh. It can be found 

along the Snowy Mountains Highway, Fastigata 

Road and Packers Swamp Road, and probably 

other forestry roads in the vicinity. 

141


• Myoporum bateae, a small pink-flowered shrub 

which also has a restricted distribution, on the 

coastal strip between Nowra and about Cobargo. 

It occurs on roadsides and in adjacent forested 

privately owned land between the Murrah and 

Tilba Tilba. Only part of this area is within Bega 

Valley Shire. 

Two other species that are of interest occur on 

Sapphire Coast Drive verges. 

• Astrotricha sp. aff. ledifolia occurs in New South 

Wales only in the Bournda and Tura Beach area, 

and in Victoria at a single location in the Upper 

Snowy River area. Despite this restricted 

distribution it has not been listed as threatened 

because a population is known within Bournda 

National Park (K McDougall, pers. comm.). It is 

common, and when flowering in spring, 

conspicuous along the edge of the mown verge 

and on cuttings on Sapphire Coast Drive 

between Widgeram Road and the Merimbula 

rubbish tip entrance. 

• Grevillea mucronulata, a small shrub with green 

flowers, occurs in the same area, although being a 

smaller and less conspicuous plant, its abundance 

on road verges is less easy to determine. This 

species occurs on the coast from the Hunter 

valley south to ‘about Eden’, but the population 

known from the Tura Beach area is of interest 

due to being widely separated (disjunct) from 

other known populations (B Makinson, pers. 

comm.). 

2. Regionally significant plant species 

Keith (1996) lists 32 plant species which are known 

in Bega Valley Shire only from remnant vegetation 

on private lands. Since the publication of that report 

several more species have been recorded from the 

South Coast for the first time, mostly from private 

lands or roadsides within agricultural areas of the 

Bega or Towamba valleys. In this context ‘South 

Coast’ refers to the South Coast botanical division 

which includes the area from Nowra to the Victorian 

border, not just Bega Valley Shire. 

Species which are known solely or principally from 

roadsides or other public reserves such as cemeteries 

are: 

142 

• Bulbine bulbosa – this yellow flowered lily had 

not been recorded by the Sydney Royal Botanic 

Gardens herbarium from the south coast prior to 

1996, when it was found on Upper Cobargo 

Road, Brogo. It appears to flower only in wet 

springs, remaining dormant as bulbs in some 

years. Mass flowerings can be quite conspicuous. 

It is now known to occur at a number of roadside 

locations as well as in Towamba and Bemboka 

cemeteries. This species has a wide distribution 

in eastern Australia in grasslands and grassy 

woodlands, and can be locally abundant, but it 

persists only in relatively undisturbed sites where 

livestock grazing pressure is low or absent. 

At all Bega valley roadside locations the plants 

are conspicuous by their absence from the 

adjacent paddock. 

• Jacksonia scoparia – this two- to three-metre 

leafless shrub is apparently also sensitive to 

grazing, since it occurs primarily on roadsides, 

and is only found in paddocks where steep banks 

limit stock access. Main occurrence is in the drier 

parts of the Bega valley, from Bemboka to 

Tantawangalo and east to about Bega. Its main 

distribution in New South Wales occurs north of 

Nowra, with the Bega valley populations being 

disjunct (distant from other known 

populations). 

• Velleia paradoxa – this yellow-flowered perennial 

herb occurs widely, though not commonly, 

throughout New South Wales in sclerophyll 

forest and grassland, but had not been recorded 

from the south coast until recently. Initially it 

was known from only one occurrence on the 

Snowy Mountains Highway, at Numbugga, but 

the survey in October 1999 located another five 

populations in the Numbugga and Bemboka 

areas. All known populations are located on 

cutting faces where there is abundant bare soil 

and little competition from grasses, except on 

Sams Corner Road, Bemboka, where roadside 

slashing has kept grass density low. 

• Leptorhynchos squamatus – this small, 

yellow-flowered daisy has a wide distribution in 

New South Wales and other southern States, 

‘usually at higher altitudes, in open areas or 

low-growing grassland’ (Harden 1992). It is


reasonably common in remnant grassland on the 

southern tablelands and in East Gippsland but 

had not been recorded from the south coast until 

recently, when it was found near the southern 

end of New Buildings Road. The October 1999 

survey located another two populations on the 

verges of New Buildings Road, one of which is 

contiguous with a large population in an 

adjacent paddock. It is also present in the 

Wyndham cemetery. 

• Leucochrysum albicans ssp. albicans – this 

conspicuous yellow-flowered perennial 

everlasting daisy has a similar wide distribution 

which was thought to exclude the south coast. 

Only one population has been found so far, on 

the Bega to Candelo road, growing on a bare 

cutting face and on top of the cutting amongst 

sparse kangaroo grass (Themeda australis). The 

plants are quite numerous. Several other 

significant herb species are also present. 

• Cullen microcephalum (Syn. Psoralea adscendens) 

– this herb in the pea family has conspicuous 

heads of purple flowers and wrinkled clover-like 

leaves. Although it occurs very close to the 

western boundary of Bega Valley Shire on top of 

the escarpment in grassy snow gum woodlands, 

it had not been recorded from the coast until its 

recent discovery in Bemboka cemetery. Outside 

this area it occurs on the Southern Tablelands 

and adjacent areas of Victoria, and in Tasmania. 

The conservation significance of the above species 

(excluding the shrub Jacksonia) in the Bega valley 

arises through them being species which are typically 

found in native grasslands or grassy woodlands. 

Their habitat has therefore been much reduced by 

agricultural activities throughout New South Wales, 

including the Bega valley. While there has been 

reasonably extensive sampling of this habitat within 

the Bega valley (Keith 1994) these species were not 

detected, so it must be assumed that they are rare in 

the district. 

There are numerous other species listed by Keith 

(1996) which are thought to be confined to the types 

of remnant vegetation now found primarily on 

private land. Many of these would occur on 

roadsides, but they have not been discussed here 

because they have been found reasonably frequently 

on private property, and so are not thought to be 

largely confined to roadsides as the above species are. 

However, their occurrence on roadsides might 

render that area of roadside significant, particularly if 

several such species were present together. These 

species are listed at the end of this report. Many are 

also grasses or herbs typical of the understorey of 

grassy open forests. 

• Eucalyptus spp. – some eucalypts are of regional 

conservation significance, either because they are 

uncommon in the district or because they are 

regarded as depleted in the wild, (that is, their 

habitat has been largely cleared for agriculture), 

or both. Two species, forest red gum (Eucalyptus 

tereticornis) and blue box (E. baueriana) are 

depleted in the wild, although they are not 

uncommon in the area. However, their 

long-term existence in agricultural landscapes is 

threatened by dieback and a lack of regeneration 

caused by livestock and possibly insect grazing 

pressure. Remnant individual trees and 

regeneration are therefore significant and should 

be retained. 

Three species which are more typical of the 

tablelands and not usually found at coastal 

elevations are found uncommonly in the district. 

They are yellow box (E. melliodora), snow gum 

(E. pauciflora) and broad-leaved peppermint 

(E. dives). Yellow box occurs mainly in the area 

from Bemboka south to Tantawangalo, and not 

much east of Kameruka, as well as at the western 

end of the Towamba valley at Rocky Hall. Snow 

gum occurs around Candelo, east to Wolumla 

and almost to Bega, in the western end of the 

Cobargo valley at Yowrie and on Murrabrine 

Lane, and along New Buildings Road west of 

Wyndham. Broad-leaved peppermint seems 

quite common at Springvale and is scattered in 

other areas, such as Frogs Hollow, Toothdale and 

Numbugga. Any individuals of these species 

occurring on roadsides should be retained, along 

with any regeneration. 

Two other eucalypts are potentially of 

conservation significance in the area, although 

they are probably more common within public 

lands (State Forests and reserves) than in 

agricultural areas. They are swamp gum 

143


(E. ovata) and red box (E. polyanthemos ssp. 

vestita). Swamp gum occurs mainly at higher 

altitudes such as at Nunnock Swamp on Packers 

Swamp Road, and in the hinterland south of 

Eden. However, it also occurs on private 

property in drainage lines especially around 

Kalaru, Frogs Hollow to Wolumla, and 

Wyndham. Much of its habitat has been 

deliberately drained for agriculture, or 

inadvertently drained through gully erosion, so 

that it is now uncommon. Main roadside 

occurrences are on the Tathra Road between 

Kalaru and the northern end of Sapphire Coast 

Drive, on Old Wallagoot Road from the 

outskirts of Kalaru south to Penuca Road, on the 

Candelo–Wolumla road west of the Wolumla 

Creek bridge and on Burragate Road near its 

junction with the Mount Darragh Road. 

Red box (E. polyanthemos ssp. vestita) is 

principally a Victorian species, with the Bega 

area being a local northern distribution limit. It 

is not uncommon in some of the hinterland 

forests south of Eden but is not well reserved, 

being found mostly in production forests. It also 

occurs in the Towamba valley where it is not easy 

to distinguish from the co-occurring blue box, at 

the northern end of Black Range just south of 

Bega, and on the adjacent western edges of Jellat 

flats. Roadside trees are known from these 

locations. 

3. Representative vegetation types 

Keith and Bedward (1998, 1999) have defined four 

forest types, described below, which once occurred 

over broad areas of the Bega Valley Shire, but which 

have now largely been cleared or substantially 

degraded by agricultural activity. They are generally 

found on the better soils in the district, derived from 

granitic rocks or various volcanic rocks, and on 

undulating to hilly rather than steep lands. None of 

them are represented in reserves to any significant 

extent. 

Bega Wet Shrub Forest (type 19) is characterised by 

the trees river peppermint (E. elata), blue box and 

black wattle (Acacia mearnsii), sometimes also 

including apple-topped box (E. angophoroides), 

ribbon gum (E. viminalis) and narrow-leaved 

peppermint (E. radiata). The most extensive stands 

144 

in near-natural condition occur along Burragate 

Road, between Wyndham and Burragate. Roadside 

stands in this area are significant because they are 

largely unaffected by clearing and weed invasion, 

compared with those in the Bega and Towamba 

valleys. This forest type also occurs at the eastern end 

of the Towamba valley, where Ben Boyd Road and 

The Snake Track pass through the coastal range. 

Other significant occurrences are along Wallagoot 

Lane on the eastern side of Jellat flats, along the 

Yowrie and Wandella roads and the western half of 

Upper Brogo Road. 

Apart from these sites the type is generally 

represented by scattered individuals or clumps of 

blue box and river peppermint, for example on 

South Wolumla Road, Old Soldier’s Road, Wanatta 

Lane. 

Brogo Wet Vine Forest (type 18) generally occurs 

on steeper sites on granite. Characteristic tree species 

are forest red gum, coast grey box (E. bosistoana), 

Port Jackson fig (Ficus rubiginosa), hickory wattle 

(Acacia implexa) and kurrajong (Brachychiton 

populneus). Known roadside occurrences are on the 

Snowy Mountains Highway at Numbugga, about 

one kilometre south-east of Garfield Road, where 

there are several figs on the roadside, and on Bald 

Hills Road west of Pambula. 

Candelo Dry Grass Forest (type 21) occurs in the 

drier rain shadow parts of the Bega valley in the area 

between Bemboka, Tantawangalo and Kameruka. 

Characteristic trees are forest red gum, rough-barked 

apple (Angophora floribunda), yellow box, hickory 

wattle and occasionally snow gum, and the 

understorey is generally grassy, potentially carrying a 

wide range of grasses and herbs, though generally 

dominated by kangaroo grass (Themeda australis) or 

weeping grass (Microlaena stipoides). Occurrences on 

roadsides include the Snowy Mountains Highway 

between Bemboka and Numbugga, the 

Bega–Candelo road and numerous minor roads such 

as Kameruka Road, Wyndham Lane, and Slaters 

Lane. There are numerous areas along rural roads in 

the central part of the Bega valley where the 

vegetation is dominated by kangaroo grass, but lacks 

trees. These areas could also be regarded as remnants 

of Candelo Dry Grass Forest, and are potentially of 

conservation significance. All of the individual


species referred to above in ‘2. Regionally significant 

plant species’, are often or only associated with this 

forest type. Unfortunately occurrences are generally 

small and interspersed with areas dominated by 

exotic grasses and agricultural weeds. 

Small occurrences of Candelo Dry Grass Forest have 

been modelled as occurring in the Towamba valley, 

despite the fact that the principal tree species, forest 

red gum and angophora, do not occur in this area. 

However, at least in some mapped locations a similar 

grass–herb understorey occurs, so these locations 

should probably be regarded as being Candelo Dry 

Grass Forest and hence of significance. One such 

location is on New Buildings Road, where snow 

gum, the herb Leptorhynchos squamatus discussed 

above, and several other herbs characteristic of 

type 21 vegetation occur. From the occurrence of a 

similar grass–herb assemblage on cuttings around 

Wyndham, it appears that this vegetation type may 

have been more extensive around Wyndham than 

modelling predicts. Roadsides at Rocky Hall carry 

yellow box, and should also be regarded as belonging 

to this vegetation type. 

Bega Dry Grass Forest (type 20) is characterised by 

forest red gum, angophora and black wattle, with an 

understorey including the shrub blackthorn 

(Bursaria spinosa) and diverse grasses and herbs. It 

occurs in agricultural areas within the Bega valley 

and Cobargo–Quaama area not carrying one of the 

other three forest types described above. Significant 

roadside remnants occur at Frogs Hollow between 

the Princes Highway and the old highway, along 

Wanatta Lane and on Coopers Gully Road. There 

are coastal occurrences which approximate this 

vegetation type, such as around Tanja on soils 

derived from syenite. 

In general, any vegetation containing eucalypts (or 

angophora) occurring on roadsides or other public 

reserves within the Towamba, Bega or Cobargo 

agricultural areas should be regarded as potentially of 

moderate to high conservation significance, since it 

is very likely to belong to one of the four vegetation 

types described above. Degree of significance will 

depend upon its condition. Factors influencing 

significance were outlined in Part B, section 7.1. 

One respect in which roadside vegetation differs 

from remnant vegetation on private property is that 

stands consisting only of regenerating trees should 

not be regarded as being of lower significance. It is a 

common occurrence for relict trees in paddocks to 

produce regeneration only on the adjacent roadside 

(due to stock grazing pressure in the paddocks), and 

this regeneration represents the future supply of trees 

in this area, once the original relict trees have died. 

4. Fauna habitat 

In general, roadside remnants are not likely to be 

highly significant as fauna habitat in Bega Valley 

Shire. This is because they are generally small and 

fragmented, often lack mature trees, and are smaller 

and less common than remnants occurring on 

marginal farming lands. Nonetheless, they can make 

a contribution to the maintenance of populations of 

farmland fauna such as birds, some mammals and 

reptiles by providing foraging habitat, shelter and 

nesting sites. 

There are instances where particular roadside trees 

are of greater significance to fauna. An obvious 

example is that of the large hollow-bearing tree 

which provides nesting or den sites for birds and 

other fauna such as possums and insectivorous bats. 

One instance where roadside trees may be significant 

for fauna, even where adjacent land is forested, is 

that of the yellow-bellied glider sap feed tree. 

Yellow-bellied gliders, listed as vulnerable in the 

Threatened Species Conservation Act 1995, chew 

through bark of eucalypt species to stimulate a sap 

flow on which they then feed. At some times of year 

sap can form a substantial part of the diet. They are 

quite selective about which trees they feed on, with 

favoured trees receiving very heavy use and other 

trees none. Favoured trees are therefore quite 

important for the maintenance of local populations, 

and they may occur on roadsides. Known examples 

are: 

• on Burragate Road near the Wyndham tip 

(Eucalyptus viminalis and E. angophoroides); 

• on Back Creek Road at Lochiel 6.9 kilometres 

south of Mount Darragh Road (bloodwood, 

Corymbia gummifera); 

• at Yowaka on Hardakers Road about 50 metres 

in from the intersection with the Princes 

145


Highway (bloodwood, Corymbia gummifera); 

and 

• on the highway about 500 metres east of 

Hardakers Road (bloodwood, Corymbia 

gummifera). 

5. Habitat connectivity 

An assessment of the degree to which roadside 

vegetation provides corridors for wildlife movement 

would require a study of aerial photos, which has not 

been done. In general, because of the small and 

fragmented nature of existing roadside remnants in 

agricultural areas, they are unlikely to contribute 

much to wildlife corridors at present. However there 

is considerable potential for improvement in this 

area if strategic plantings were made, or regeneration 

encouraged in some areas. A change in Council 

policy with regard to roadside slashing in the last 10 

years or so has already seen considerable amounts of 

tree regeneration spring up along some areas of 

roadside, where a suitable seed source exists. This 

will in time provide a good resource for wildlife, as 

well as shade trees for the owners of adjoining 

paddocks where regeneration is prevented by stock 

grazing pressure. 

Roadside vegetation may also be significant in 

forested areas where it contributes to wildlife 

corridors. The presence of a road may provide a 

significant impediment to movement of some types 

of fauna, but this can be minimised by permitting 

trees to remain on the verges, particularly if their 

canopies actually meet above the road. This would 

make movement of arboreal fauna, such as possums 

and gliders, easier and safer. Examples of locations 

where roadside vegetation could contribute 

significantly to wildlife corridors are at the southern 

end of Wallagoot Lane west of Kalaru racecourse, 

and some sections along Sapphire Coast Drive 

between Bournda and Merimbula. 

6. Recommendations for roadside 

management 

The following are some general recommendations 

for roadside management in Bega Valley Shire. 

• Maintain Council’s current minimal slashing 

policy on most road verges to encourage tree 

regeneration. However, where significant grassy 

146 

remnant vegetation with a high diversity of 

native herbs occurs, the herbs are generally able 

to persist only because past management has 

involved regular reduction of kangaroo grass 

volume, either by slashing or burning. It will be 

necessary to plan and implement a management 

regime for these sites as it is possible that without 

such management the native herbs will 

eventually be smothered by grass growth and 

disappear from these sites. 

• Restrict lopping crews to removing senescent 

black wattle where it could fall on the road, and 

other vegetation only where it definitely 

obstructs visibility. If lopped material is stacked 

for burning rather than chipped, ensure that this 

occurs on disturbed verges, away from trees and 

tree or shrub regeneration and away from 

identified significant sites. 

• Appraise ground crews of the currently known 

locations of significant plant species and 

management guidelines for those sites. In general 

they should avoid damaging any mature native 

trees or regeneration and avoid disturbing areas 

obviously carrying native vegetation (including 

treeless areas dominated by kangaroo grass), 

except in accordance with management plans for 

significant sites. 

• Weed spraying crews should be trained in 

recognition of the target species. A number of 

significant herb species have yellow flowers, and 

could be sprayed in mistake for fireweed or 

St John’s wort by operators with a fairly broad 

search image. 

• Where gravel dumps or similar sites are needed, 

they should be located on places which carry 

only exotic vegetation, although sites carrying 

noxious weeds, such as African lovegrass, should 

be avoided. 

• If material needs to be won from road verges for 

re-surfacing unsealed roads, it should be taken 

from sites which carry only exotic vegetation. 

Cutting faces are often good sites for the 

retention of largely native vegetation, since the 

harsh growing conditions on them tend to 

exclude exotic plants from establishing there. 

Disturbance of these sites to obtain road base


Roadside species 

Listed below are species of regional conservation significance which could occur on roadsides in Bega 

valley agricultural areas, but are not confined to them. 

Trees 

Eucalyptus baueriana E. dives 

E. melliodora E. ovata 

E. pauciflora E. polyanthemos ssp. vestita 

E. tereticornis Ficus rubiginosa 

Shrubs 

Jacksonia scoparia 

Grasses and sedges 

Aristida ramosa var. ramosa Austrodanthonia spp. 

Austrostipa scabra ssp. falcata Austrostipa verticillata 

Bothriochloa macra Chloris truncata 

Digitaria ramularis Elymus scaber 

Sorghum leiocladum 

Sporobolus elongatus 

Sporobolus creber 

Cyperus gracilis Scleria mackaviensis 

Herbs 

Brachycome ciliaris var. ciliaris Calotis lappulacea 

Chrysocephalum apiculatum Chrysocephalum semipapposum 

Convolvulus erubescens Desmodium brachypodum 

Glycine tomentella sens. lat. Lespedeza juncea 

Linum marginale Pimelea curviflora var. sericea 

Polygala japonica 

Zornia dyctiocarpa var. dyctiocarpa 

Tricoryne elatior 

In general, the presence of the easily recognisable kangaroo grass (Themeda australis) is a reasonable 

indicator of the possible presence of most of the above species. 

may therefore present problems. If material is 

taken from sites with exotic vegetation, the 

topsoil should remain on the site, rather than 

being spread with its burden of weed seeds onto 

other sites. Weed invasion is a major threat to 

roadside remnant vegetation. The establishment 

of a quarry to supply road base material within 

the Shire should be a high priority, to prevent ad 

hoc movement of soil around the roadsides of 

the Shire. Weed control at such a quarry would 

need to be rigorously applied to prevent the 

spread of weeds around the Shire. 

• When undertaking slashing on road verges, work 

from clean areas to more weed-infested areas 

rather than the reverse, to avoid spreading weeds 

onto areas that are currently free of them. This 

would be difficult to implement on a small scale, 

since native vegetation and weeds alternate in 

rapid succession along most roads in agricultural 

areas. On the broader scale it would involve 

leaving core areas of infestation of African 

lovegrass, serrated tussock, St John’s wort and 

fireweed till last when slashing, and washing 

machinery thoroughly before re-use. Washing 

147


should be done in an area where run-off can be 

retained on site and weed growth can be 

monitored and controlled. 

• Complete the survey of roadsides in the Shire to 

detect any further sites of high significance. This 

would need to be done by observers with a high 

degree of botanical knowledge, since the 

difference between a native grass or herb site and 

an exotic one can be too subtle for detection by 

untrained observers. If such a survey is done in 

stages, then the next stage should concentrate on 

the agricultural areas north of the Brogo Pass, 

which were omitted from the October 1999 

survey. The survey should preferably proceed in 

tandem with one of other public lands such as 

travelling stock reserves, vacant crown land, 

showgrounds and cemeteries. The best timing 

for such a survey would be late spring and 

summer. 

• There may be an advantage in having the 

location of all remnant vegetation (trees and 

kangaroo grass-dominated sites) mapped on 

Council’s geographic information system. The 

present work has not resulted in mapping of such 

vegetation, only of that containing significant 

148 

herb species and the regionally uncommon 

eucalypts (yellow box, snow gum and 

broad-leaved peppermint). If such mapping is 

considered desirable, it could probably be done 

by council staff with a minimal amount of 

training. 

• Require groups or individuals who want to do 

roadside plantings to use appropriate, local 

native species from locally-sourced seed. There 

are numerous plantings on highway verges, 

particularly around Bemboka and Wolumla, 

which consist of non-local native trees and 

shrubs. The Roads and Traffic Authority has 

been responsible for some of these, but 

community groups and individuals have 

presumably also been involved. Such plantings 

have the potential to reproduce themselves, 

spreading beyond the planting site and possibly 

displacing local natives. Planting of locally 

occurring species from non-local seed has the 

potential to ‘contaminate’ local gene pools. 

Where a seed source is available from existing 

roadside vegetation it would be preferable to 

encourage natural regeneration rather than to 

plant.

Vegetation types described for Bega Valley Shire - NSW Government

Create successful ePaper yourself

Delete template?

Save as template?