Adrienne markey

Yüklə 7,96 Mb.
Pdf görüntüsü
ölçüsü7,96 Mb.
  1   2   3   4   5   6   7   8   9   ...   14

Conservation Science W. Aust. 8 (2) : 187–239 (2012)

© The Government of Western Australia, 2012

Floristic communities of the Ravensthorpe Range,

Western Australia







Science Division, Department of Environment and Conservation, Locked Bag 104, Bentley Delivery Centre,

Western Australia 6983. Email:;

Western Botanical, PO Box 3393, Bassendean, WA, 6842.

current contact email:


The Ravensthorpe Range consists of a series of ridges and peaks located in the greater Fitzgerald Biosphere reserve, in

south-west Western Australia near the township of Ravensthorpe. A floristic classification of the vegetation of the

range is described and compared to recent vegetation mapping undertaken in the area. The floristic classification was

derived from an analysis of 580 taxa from 265 plots that were located to cover the topographic and geographic extent

of the range. Twenty-one communities are described, and these show broad agreement with the recent detailed mapping,

except for one widespread upland unit that was floristically heterogeneous. The floristic communities correlated with

site physical attributes, notably topographic position, substrate and altitude. In addition to the high beta diversity, the

Ravensthorpe Range has a large number of geographically restricted species, species listed as threatened and species

being considered for listing. Despite the high conservation values of the range, reservation is limited to two small A-

class nature reserves off the main range. These reserves are not representative of the full diversity of flora and communities

on the range. Mineral exploration and mining are currently active or proposed for parts of the range, and mining

tenements cover most of the area of the range. The impacts of exploration, mining and Phytophthora present significant

challenges to managing this major biodiversity hotspot.

Keywords: endemic, threatened taxa, beta diversity, classification


The Yilgarn Craton of Western Australia is composed of

ancient granitoid gneisses, within which are inter-bedded

extensive belts of metamorphosed sedimentary and

volcanic rocks. These are referred to locally as ‘greenstone’

belts, which manifest as hills and ranges when exposed

above the surrounding plains. Recent surveys of these

landforms have found them to be repositories of endemic

and rare plant taxa (Gibson et al. 2007, 2010, 2011), and

short range endemic invertebrate species (i.e. species with

total range of less than 10,000 km


; Harvey 2002; Edward

& Harvey 2010). They also support floristic communities

which differ in composition to both the surrounding

lowlands and adjacent ranges (Gibson et al. 2007, 2010,


The Ravensthorpe Range is one such greenstone

landform located on the Yilgarn Craton, in the far south-

western corner of Western Australia (Fig. 1) and within

the South West Australian Floristic Region (SWAFR,

sensu Hopper & Gioia 2004). This region is known for

its exceptional biological diversity, and is considered to be

one of 34 global biodiversity hotspots because of a

combination of high species richness, endemism and

habitat loss (Myers et al. 2000; Myers 2003). The

Ravensthorpe Range occurs within a node of high species

diversity and endemism within the SWAFR (Hopper &

Gioia 2004), and is part of the United Nations

Educational, Scientific and Cultural Organisation’s

Fitzgerald River Biosphere Reserve (Chapman & Newbey

1995). Because of its location, the Ravensthorpe Range

is recognised as a significant biological corridor between

the Fitzgerald River National Park and the Great Western

Woodlands (Chapman & Newbey 1995; Watson 1991;

Watson & Wilkins 1999). The range is noted for its high

biodiversity values (Craig et al. 2008; Moir et al. 2009;

Edward & Harvey 2010; McDonald 2010), including

being important habitat for several threatened species of

vertebrates (Chapman & Newbey 1995; Harris et al.

2008). The most recent census of vascular plants recorded

1324 native species (Craig 2008) from Beard’s

Ravensthorpe vegetation system (Beard 1972, 1973),

which covers both the range and areas of mafic geology

immediately to the south-west. At least 58 taxa are endemic

to, or largely restricted to, this vegetation system (Craig

2008; Kern et al. 2008; the Western Australian Herbarium

1998–). Some 76 taxa in the Ravensthorpe vegetation

system are listed as being of conservation significance


A Markey et al.

Figure 1. Ravensthorpe Range survey area showing major landmarks and the locations of the 266 permanent plots. Darker

shading indicates upland areas.

(Craig 2008; Smith 2010), and 60 of these are known to

occur on the range (Harris et al. 2008; Western Australian

Herbarium 1998–). Five ecological communities of

conservation significance are listed for the range (Craig et

al. 2008; Kern et al. 2008; Department of Environment

and Conservation 2010). Like many greenstone ranges

on the Yilgarn Craton, the Ravensthorpe Range is rich in

mineral resources and of interest to the resources sector

(Witt 1998).


The Ravensthorpe Range is an extensive outcropping of

greenstone bedrock located 550 km south-east of Perth,

Western Australia. The main range is a narrow, linear

feature c. 50 km in length trending north-west to south-

east, and up to 7 km wide. The centre of the range is 7

km east of the town of Ravensthorpe and the southern

edge 25 km from the coastal township of Hopetoun. There

are six peaks named on the main range; Mt Short is the

highest peak (453 m), followed by Mt Benson (404 m),

Mt Desmond (340 m), Mt Chester (300m), Mt McMahon

(274 m) and Overshot Hill (377 m; Fig. 1). Immediately

east of the main ridges are a series of low, undulating

pediments and hills. Isolated from the main range,

Bandalup Hill (228 m) is a prominent feature 34 km east

of Ravensthorpe township and on the eastern margin of

the survey area (Fig. 1). The main range is relatively low

in elevation, being only, at the most, c. 150 m above the

surrounding plains. Numerous drainage lines and several

ephemeral creeks drain from the ranges into the Steere

and Jerdacuttup Rivers, which incise the central and

southern portion of the range. These two main rivers are

saline and flow only after winter rainfall, contracting to a

series of permanent pools during the summer months.


The Ravensthorpe Range is located near the edge of the

transitional rainfall zone of the SWAFR, and the area has

a warm Mediterranean climate, with cool and wet winters

and warm to hot summers (Harris et al. 2008; Beard

1990). The climate is moderated by proximity to the

Floristic communities Ravensthorpe Range


Southern Ocean, and rainfall is predominately derived

from southerly fronts that occur during the winter months

(Craig et al. 2008; Chapman & Newbey 1995). However,

summer rainfall can originate from the north as remnants

of tropical cyclones. On very rare occasions snowfalls have

been reported on the range (Chapman & Newbey 1995).

The nearest meteorological station is Ravensthorpe

township (1901–2007), which has an annual average

rainfall of 426 mm (Bureau of Meteorology 1908–). Data

from average annual rainfall grids indicate that rainfall

across the range declines from 523 to 391 mm with

distance from the coastline (Harris et al. 2008).


The Ravensthorpe Range is essentially a linear outcropping

of a complex of Archean (c. 3 Ga) metavolcanic (mafic

and ultramafic) and metasedimentary bedrock, which is

inter-bedded within the granitoid gneiss of the southern

Yilgarn Craton (Witt 1997, 1998). These metasediments

and metavolcanics are faulted and folded to produce a

sharp ridge and peaks, the uplands of which are covered

by an undulating upland of Quatenary laterites or silcretes,

produced from in situ weathering. The eroding margins

of this caprock give way to colluvial slopes with some

outcropping bedrock and deep valleys incised into steep

slopes. Surrounding the Ravensthorpe Range are Tertiary

sediments, colluvium and sand that overlie the Archean

granitoid gneiss, forming extensive, low-lying plains and

alluvial flats interrupted by outcrops and hills. The nearest

significant granite–quartzite hills are associated with the

Barren and Eyre Ranges, the closest peaks being 25–40

km south of Ravensthorpe.

The geology of the Ravensthorpe greenstone belt and

associated Ravensthorpe Range has been described and

mapped at a scale of 1:100,000 (Witt 1996). Witt (1996,

1997) described three geological terranes that cover the

Ravensthorpe Range survey area, of which the Carlingup

Terrane is the main tectonic unit. The Ravensthorpe

Terrane abuts onto the western margin of the Carlingup

Terrane, and the Cocanarup greenstones underlie part of

Mt Short. The Carlingup Terrane is composed of four

main stratigraphic units: the Bandalup Ultramafics, the

Chester Formation, Maydon Basalt and the Hatfield

Formation. The Bandalup Ultramafics (metamorphosed

undivided ultramafic rocks) occur on lower slopes and

undulating low hills, while the Maydon Basalt

(metamorphosed undivided mafic rocks) dominates low

hills between the main range and Bandalup Hill. The

Hatfield Formation (pelite, psammite, metamorphic and

felsic volcanic rocks) lies on the eastern margin of main

ridge. It is the durable, erosion-resistant Chester

Formation (pelite, psammite and metamorphosed

sedimentary rocks) which forms the main ridges of the

range (Witt 1996, 1997).

Soils of the Ravensthorpe Range reflect underlying

lithology and topography, with skeletal to shallow soils

being formed in situ on upland laterites and bedrock,

colluvium being deposited on hill slopes and foot slopes,

and alluvial deposits accumulating in drainage features and

alluvial flats at the base of the range. Associated with the

mafic and ultramafic lithologies are mineral-rich clays

(Beard 1973; Bennett 1987; Chapman & Newbey 1995).


The Ravensthorpe Range is situated within the Esperance

Plains Bioregion (Department of Environment and Water

Resources 2007). The vegetation was first described by

Beard  (1972, 1973) as part of a wider survey of the

Ravensthorpe region. According to these publications,

three vegetation systems occur on the range

(Ravensthorpe, Oldfield and Esperance Systems). The

Ravensthorpe System covers most of the range and wider

surrounds to the south-west and east. Beard recognised

eight physiognomic communities on the range. These

communities are associated with the topographic and soil

catenary sequence on the range and fall into three

categories: ‘Thickets’ on summit ridges; ‘Mallee’ on

pediments and low hills; and ‘Sclerophyll Woodland’ in

deep, broad valleys. Bennett (1987) and Chapman and

Newbey (1995) subsequently focussed on sections of the

range, providing greater floristic and structural detail for

these sections and describing their associations with

edaphic and topographic factors. Targeted surveys on other

parts of the range have been undertaken for environment

impact assessments (e.g. Cockerton & Craig 2000; Craig

1999, 2004, 2005). More recently, detailed vegetation

mapping of the main range between Mt Short and Kundip

has described 70 vegetation units based on structure and

composition (Craig et al. 2008).

Land use

Over the past century the Ravensthorpe region has been

an agricultural and mining centre, and these activities have

had impacts upon the vegetation of the range. The region

lies in the south-east of the wheatbelt agricultural zone,

and a significant proportion of the woodlands around the

range have been cleared or cut-over. Wheatfields abut the

base of the northern and western foothills, taking

advantage of the heavy, rich soils derived from the mafic

bedrock (Beard 1973).

The Ravensthorpe greenstone belt is highly prospective

for minerals, and has a long history of exploration and

mining for base and precious minerals, notably gold,

copper, nickel and silver (Witt 1998). The original mining

centre was established in the 1890s at the now abandoned

township of Kundip, and several open pits and tailings

dumps remain in the southern section of the range from

these and more recent mining activities (Harris et al.

2008). Major operations to exploit significant nickel

laterite and magnesite deposits at Bandalup Hill

commenced in January 2008, and a lithium/tantalum mine

immediately north of Ravensthorpe township (Mt Cattlin)

was approved in 2009. Exploration activities are in

progress at various locations over the range, and active or

pending mining tenements cover almost the entire extent

(Harris et al. 2008). Although not in the tenure of the

Department of Environment and Conservation (DEC),


A Markey et al.

the Department manages these areas and conducts

prescribed burns and feral animal baiting programs. Most

of the range is Unallocated Crown Land, with the

Overshot Hill and Kundip Nature Reserves being the only

parts of the range within the conservation estate (both A-

class nature reserves; Fig. 1).


The aim of this study was to produce a classification of

the vegetation communities on the Ravensthorpe Range

based on floristic composition, in order to allow the

regional context of proposed developments to be

determined in a consistent and repeatable manner. These

data were also used to examine compositional

heterogeneity in some of the 70 mapping units that cover

the northern sections of the range (Craig et al. 2008) and

to assess the relationship between the vegetation mapping

and the floristic classification.


Two hundred permanently marked, georeferenced plots

were established by a team of consultants (Western

Botanical Pty. Ltd.) and rescored over a period of several

months in 2007. A further 66 plots were established by the

DEC in the following late spring and summer of 2008–09

(Fig. 1). Plots were placed to cover the geographical and

topographical extent of the range in order to capture the

widest range of floristic variation. No recently burnt

vegetation was sampled in order to minimise the inclusion

of fire ephemerals in the dataset. For each plot, location

(to within an accuracy of 4 m) and altitude were

determined using a GPS (Garmin, GPS76). The

occurrence of all vascular species was recorded in each 10

× 10 m plot. Each of these plots was nested within a larger


× 20 m




that was used to record canopy species

(trees, mallee shrubs >3 m). Plot size was based on species-

area curves from Bennett (1987) and sampling

methodology follows Gibson et al. (2004). Taxa occurring

within and adjacent to plots were collected for

identification at the Western Australian Herbarium

(PERTH), where over 1500 voucher specimens have been

lodged. Further information on taxa was obtained from

online records of the Western Australian Herbarium


Data analyses were performed using routines available

in P


-E and P


+ (P


 version 6, Clarke

& Gorley 2006). Presence/absence data were analysed

using Bray & Curtis association measures. This measure

has proved extremely robust in the analysis of ecological

data (Faith et al. 1987). Classifications of both the plots

and species were performed using UPGMA clustering

(Sneath & Sokal 1973). Similarity profiles (SIMPROF)

and indicator species analysis (INDVAL) were used to

examine the classification structure (Dufrene & Legendre

1997; Clarke et al. 2008). Floristic nomenclature follows

that of the Western Australian Herbarium (1998–).

A preliminary analysis was undertaken on 593 taxa

identified from within the 266 plots. The subsequent

analysis was performed on 580 taxa from 265 plots, after

one outlying plot and its singleton taxa were omitted.

Hybrid taxa were also omitted and several closely-related

or intergrading pairs of taxa were amalgamated at a higher

taxonomic level, mostly from lower ranks to species level

(Hybanthus floribundus, Hakea pandanicarpa, Eucalyptus

astringens, Melaleuca pentagona, Acacia sulcata, Pultenaea

indira and Schoenus sublaxus) or a species complex later

resolved to be separate taxa (Leucopogon sp. Coujinup

[MA Burgman 1085] and Leucopogon sp. Newdegate

[M Hislop 3585], and Lysinema ciliatum and Lysinema

pentapetalum). Preliminary analyses showed that these

amalgamations had little effect on the classifications.

For each community a list of diagnostic, common,

short-range endemic and conservation taxa was compiled

and a photograph of a plot representative of each

community is provided. These photos were taken along

plot diagonals. Diagnostic taxa indicative of a community

type were identified using indicator species analysis

(INDVAL), implemented in PC-ORD (Dufrêne &

Legendre 1997; McCune & Mefford 1999). Indicator

values were derived from species fidelity and constancy

within each community, where a value of 100 indicated

that a species was present in all plots of a community and

absent from all other plots. Diagnostic species were defined

as those with INDVAL values 14 or greater. Common

taxa were those that occurred at frequencies of 

≥25% of

plots in a community. Short-range endemic (SRE) taxa

were defined as those that had a total range of less than

10,000 km


. Four categories are listed: 1, taxa with ranges

<10 km


; 2, taxa with ranges 10–100 km


; 3, taxa with

ranges 100–1,000 km


; 4, taxa with ranges 1,000–10,000



. Convex hulls were calculated from distributional data

downloaded from the Western Australian Herbarium,

which include our vouchers (accessed 22nd November

2010). Conservation taxa listing follows Smith (2010)

and recent updates as shown on Florabase (Western

Australian Herbarium 1998–): R = threatened taxa

formally listed under Western Australian legislation; P1–

P4 = taxa being considered for listing (codes follow Smith


Typical soils (McDonald et al. 1998) and the number

of plots in each of the vegetation map units of Craig et al.

(2008) were also listed for each community. Vegetation

map units were determined by intersecting the vegetation

map with plot location. Some 166 of the 266 plots fell

within the mapped area; seven of these were in mosaic

units that were not included in the community

descriptions. Plots were located in 40 of the 70 vegetation

units recognized by Craig et al. (2008), with frequencies

ranging from 1 to 44 plots per vegetation unit.



A total of 697 taxa (species, subspecies and varieties,

including putative new taxa and nine new putative hybrid

Floristic communities Ravensthorpe Range


or intergrade taxa) were recorded within and adjacent to

the 266 plots (Appendix 1). Of these, six are listed as rare

and 37 are on DEC’s priority list (Appendix 1). Six species

are naturalised herbaceous weed taxa (Asphodelus

fistulosus,  Avena fatua,  Lepidium africanum, Sonchus

oleraceous, Lolium perenne and Asparagus asparagoides).

The highest species numbers were from Myrtaceae (136

taxa), Fabaceae (112; 64 Papilionoidea), Proteaceae (69),

Cyperaceae (53), Ericaceae (28) and Goodeniaceae (23).

The most species-rich genera were Eucalyptus (45 taxa

and 8 hybrids), Acacia (46), Melaleuca (35),

Lepidosperma (33), Hakea (19), Grevillea (17), Daviesia

(14 and 1 hybrid), Boronia (13), Schoenus (14) and

Hibbertia (11).

Craig (2008) listed 55 taxa believed to be endemic

(99–100% of known populations) or almost endemic (80–

99% of known populations) to Beard’s (1972, 1973)

Ravensthorpe vegetation system that covers most of the

range as well as extensive areas to the south-west. These

definitions are somewhat arbitrary and rely on boundaries

defined at 1:250,000-scale mapping. In contrast, we

defined short-range endemic taxa as those taxa recorded

in our survey that had a total range of less than 10,000



, following definitions used for invertebrate taxa

(Harvey 2002). Some very restricted taxa fall outside this

definition, for example taxa with a small fragmented

population restricted to rare habitats but with a total range

of >10,000 km


 (e.g. Tetratheca applanata—total range

10,072.5 km


). Sixty-four SREs were found, four

restricted to a range of less than 10 km


, and a further

seven with ranges of less than 100 km




These taxa were

not restricted to particular families but primarily

represented shrubs, with a minor occurrence of eucalypt

mallees, trees and mallets (Table 1).


All but two of the 265 plots were classified into 21 groups

(subsequently referred to as communities; Fig. 2). Groups

have been resolved at variable similarity levels, and

decisions on selecting groups at particular levels have been

based on branch support from SIMPER analysis, from

Yüklə 7,96 Mb.

Dostları ilə paylaş:
  1   2   3   4   5   6   7   8   9   ...   14

Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur © 2020
rəhbərliyinə müraciət

    Ana səhifə