Survey of drill
lines for CS
7 CS flora
4 CS flora
1 T (DRF)
Survey of drill
lines for CS
Table 4 lists the personnel involved in both fieldwork and plant identifications for the field
survey conducted in 2012. All personnel have had previous field experience in the Swan
Coastal Plain region, with personnel involved in plant identifications having extensive
taxonomic experience with the flora of the Swan Coastal Plain. All plant material was
collected under the scientific licences (pursuant to Wildlife Conservation Act 1950 (WC Act)
Section 23C and Section 23F) as listed in Table 4.
Alison Saligari (WEC)
SL009956 / 149-1112
Bethea Loudon (WEC)
SL009953 / 150-1112
Greg Woodman (WEC)
SL009959 / 144-1112
John Grantham (WEC)
Terri Jones (WEC)
Frank Obbens (WEC contractor)
Melissa Bestwick (WEC casual)
Melissa Hay (WEC casual)
Peter Malajczuk (WEC casual)
Sharnya Thomson (WEC contractor) Fieldwork / Plant
Alex Devine (NAC)
Jacquie Milner (NAC)
SL010034 / 35-1213 (Mid-West
Sharon Hines (NAC)
Simon Colwill (WB)
Daniel Brassington (WB)
Manon Hannart (WB)
Wesley Lamont (WB)
Note: Personnel companies are abbreviated as follows: WEC: Woodman Environmental Consulting, NAC:
Initial interpretation of vegetation boundaries was conducted with the use of ortho-rectified
aerial photography at a scale of 1:10,000, sourced by Woodman Environmental from SLIP,
with photography dated February 2012. Preliminary vegetation type (VT) boundaries were
transcribed onto the aerial photography, to allow for ground-truthing of these boundaries to be
conducted in the field. Preliminary quadrat locations were also allocated based on these
vegetation type boundaries. A minimum of three quadrats were allocated to each discernable
vegetation type where possible. Such replication is required for meaningful results to be
produced following statistical analysis of quadrat data, and to provide local context for VT
A total of 235 quadrats were established within the Study Area over six field trips totalling 25
days (65 team days) as listed below:
Field trip 1:
24th – 28th September 2012 (10 personnel)
8th – 12th October 2012 (8 personnel)
Field trip 3:
22nd – 26th October 2012 (2 personnel)
Field trip 5:
19th – 23rd November 2012 (4 personnel)
Field trip 7:
The Study Area was accessed by vehicle where possible, using available tracks and public
All quadrats covered an area of 100m
with quadrats measuring 10m x 10m. This quadrat
standard size used in South-West of Western Australia flora and vegetation surveys. Quadrats
were established in all vegetation types identified following interpretation of aerial
photography. The number of quadrats within each identified vegetation type was determined
based on the size of the area covered by the vegetation type, and the potential species richness
of each plant community. Multiple quadrats were established in all identified vegetation
types where possible.
All quadrats established were non-permanent, with measuring tapes extended to define the
with the bearings of each side recorded for any quadrats that could not be established using
cardinal points. All vascular taxa that were visually identifiable within each quadrat were
recorded and collected as necessary. The following information was recorded at each
Unique quadrat number
Date of survey
GPS coordinates (WGS84), and location at quadrat where coordinates were recorded
Topography (including landform type and aspect)
Soil colour and type (including the presence of any rock outcropping and surface
displayed in Table 5)
Approximate time since fire
Presence of disturbance (if any)
Percentage foliage cover (for each species)
Height (m) (for each species, excluding climbers/aerial shrubs)
Additional flora taxa were also recorded opportunistically via a search around the general
vicinity of each quadrat, during traverses on foot between quadrats and while driving along
Pristine or nearly so; no obvious signs of disturbance.
Vegetation structure intact, disturbance affecting individual species and weeds are non-
Vegetation structure altered, obvious signs of disturbance.
Vegetation structure significantly altered by very obvious signs of multiple disturbances.
Retains basic vegetation structure or ability to regenerate it.
Basic vegetation structure severely impacted by disturbance. Scope for regeneration but
The structure of the vegetation is no longer intact, and the area is completely or almost
Specimens of any unknown taxa were collected and pressed for later identification at the
WAHerb. Identifications were undertaken by experienced taxonomists Frank Obbens and
Sharnya Thomson, with experts in particular families or genera consulted for any specimens
considered to be of taxonomic interest. Species nomenclature follows FloraBase (DPaW
2013a) with all names checked against the current DEC Max database to ensure their validity.
The conservation status of each species was checked against FloraBase (DPaW 2013a),
which provides the most up-to-date information regarding the conservation status of flora taxa
in Western Australia.
Specimens of interest (T-DRF and Priority Flora taxa, range extensions and potential new
Priority Flora Report Forms (TPFRF) will be submitted to the DEC for all populations
recorded of T-DRF and Priority Flora taxa.
3.5 Statistical Analysis
Quadrat data was statistically analysed to aid in the determination of VTs, using similar
methods to those previously used by Woodman Environmental in the Northern Sandplains
region (Woodman Environmental 2009e; 2011b). Data from a total of 370 quadrats
established within the Study Area were included in the analysis (Figure 6):
235 quadrats established during the current survey (September – November 2012)
27 quadrats established at Falcon (October 2006) (Woodman Environmental 2006c)
24 quadrats established as part of the Falcon regional survey (October – November
quadrats only) (October 2010) (Woodman Environmental 2011)
A summary of the quadrats used in the analysis, including the quadrat code, date of
establishment, location co-ordinates and the VT in which the quadrat was grouped are
presented in Appendix F, with locations also presented in Figure 6 and Appendix T.
The Woodman Environmental datasets (of historical data) were reviewed and updated prior to
current name after consulting FloraBase (DPaW 2013a). Any odd entities resulting from
possible erroneous identifications (i.e. species that do not occur anywhere near the Study Area
in other parts of the State) were also fixed, generally by removing the species name.
In particular, Xanthorrhoea, Macrozamia and Grevillea thelemanniana/preissii entities were
and site information, including their location within the Study Area in relation to vegetation
associations/land systems and their known area of occurrence and habitat preference, were
As a result, all previous records of Macrozamia riedlei were changed to M. fraseri as M.
appearance, apart from one site on laterite in the north-east corner of Study Area where plants
were quite obviously X. drummondii and one which was changed to X. ?drummondii as it was
not easily discernible which species it was. In the case of Grevillea thelemanniana and
taxa and their subspecies, particularly in the absence of specimens. Instead the habitat of all
quadrat locations was reviewed and compared, with the various taxa grouped together in the
analysis based on the habitat/landform in which they were recorded, i.e. entities were grouped
under thelemanniana if they occurred on lower-lying winter-wet/damp flats/plains, or under
preissii if they occurred in drier limestone associated habitats.
In addition, previous records of Eremophila glabra subsp. chlorella (T-DRF) were re-
WAHerb. These entities are now known as E. glabra subsp. ?carnosa (C).
Some entities of Lepidosperma pubisquameum are also doubtful, particularly where they are
in the absence of flowering material at the time of the previous surveys. These records have
not been changed.
Classification and ordination analyses were conducted on a data matrix compiled from the
quadrat data, with introduced taxa, putative hybrids and opportunistic recordings (i.e. those
taxa recorded outside of the quadrat) excluded from the analysis. Singletons (taxa recorded
only once in the quadrat dataset) and ephemeral (short-lived) taxa including geophytes (i.e.
entities of Orchidaceae, Droseraceae, Eryngium and Wurmbea) were also excluded from the
analysis. These exclusions remove data inconsistencies resulting from differences in seasonal
conditions and time of year in which surveying is undertaken, and the ability to record such
Various taxa were grouped together within the data matrix for the analysis where taxonomy
taxa occurred and did not correlate to different plant community, landform or soil types.
Some recordswere omitted from the analysis where the taxa could not be positively identified
(e.g. Grevillea sp., ?Grevillea sp.).
Pattern analysis was conducted using PATN (V3.03) (Belbin 1989). The Bray-Curtis
ordination analyses. This association matrix consisted of pairwise coefficients of similarities
between quadrats based on floristic data. Agglomerative, hierarchical clustering, using
flexible UPGMA (ß=-0.1) was used to generate a species and quadrat classification (Sneath
and Sokal 1973). A two-way table of the species and quadrat matrix was produced, with the
matrix sorted into groups generated from the species and quadrat classification. Indicator
species analysis (INDVAL) was conducted using PC-Ord (McCune and Mefford 1999) using
the method of Dufrene and Legendre (1997). The INDVAL measures were used to determine
the indicator species for each VT and a Monte Carlo permutation test was used to test for the
significance of the indicator species.
The Study Area covers approximately 34 424 ha, with 370 quadrats established in the Study
Area; as each quadrat has an area of 0.01 ha, a total area of 3.7 ha was sampled directly by
quadrats, equating to 0.01 % of the Study Area. Quadrats were established in all VTs as
discernable by initial aerial photograph interpretation (see Section 3.2), both to adequately
sample variation in vegetation throughout the Study Area, and to ensure adequacy of
sampling for vascular plant taxa.
Species area curves detailing the number of taxa returned within quadrat data both as a whole
Ord (V 4) (McCune and Mefford 1999). The species area curves are presented in Appendix
G. The species accumulation curves were generated using all native taxa (both annual and
perennial) recorded within each quadrat, with groupings as per used in the statistical analysis
As a whole, the number of quadrats established in the Study Area satisfies the criterium of a
The number of quadrats established in VTs 1, 2, 5, 6, 7, 8, 12, 17 and 18 likewise met this
criterium (Table 1 in Appendix G); this was due to the relatively large number of samples
(quadrats) taken in each of these VTs. VTs 3, 4, 9a, 9b, 10, 11 and 13 - 16 did not meet this
criterium (Table 1 in Appendix G), invariably due to the lower number of samples taken
within these VTs. This low number is attributable to the relatively scarcity of each of these
VTs in terms of area mapped
Species accumulation calculations for the entire dataset were also undertaken using PC-Ord.
(using the Chao2 esimate (bias form corrected); and is equivalent to 78.6 % of the 95 %
confidence interval figure. Therefore it is estimated that approximately 80 % of the species
that can be expected in the Study Area were recorded within quadrats.
The species and quadrat classification generated from the statistical analysis of quadrat data
identified the floristic units (VTs). Aerial photography interpretation and field notes taken
during the survey were then used to develop VT mapping polygon boundaries over the Study
Area. These polygon boundaries were then digitised using Geographic Information System
VT descriptions (though floristic in origin) have been adapted from the National Vegetation
2003), a system of describing structural vegetation units (based on dominant taxa), preferred
by the DPaW (Dr. Stephen van Leeuwen pers. comm.). This model follows nationally-agreed
guidelines to describe and represent vegetation types, so that comparable and consistent data
is produced nation-wide. For the purposes of this report, it is considered that a VT is
equivalent to a NVIS sub-association as described in ESCAVI (2003).
After the statistical analysis, determination of vegetation types and preliminary mapping
Woodman and Bethea Loudon to determine if further refinements of the VT polygon mapping
were required. All accessible tracks were traversed as well as additional foot transects
undertaken across the Study Area to confirm VT mapping and boundaries. One site (relevé)
was undertaken during the trip to confirm the VT of a small piece of fringing vegetation in the
western portion of the Study Area. The data from this site was incorporated into the analysis.