Numerical analysis of the timbered vegetation in Tidbinbilla nature reserve, A.C.T., Australia*,** F. Ingwersen
ACT Conservation Service, Department of the Capital Territory, Canberra, ACT 2600, Australia
Keywords: Classification, Ecological survey, Eucalyptus, G r a d i e n t analysis, M a p p i n g , M i c r o c l i m a t e , N a t u r e reserve, O r d i n a t i o n , P r i n c i p a l c o - o r d i n a t e analysis
Abstract T h r e e a p p r o a c h e s to the d e s c r i p t i o n of e u c a l y p t forest v e g e t a t i o n are described. Using p r e s e n c e / a b s e n c e d a t a (81 sites, 129 species) an agglomerative, hierarchical classification m e t h o d based on i n f o r m a t i o n analysis p r o d u c e d ecologically meaningful groups. Q u a n t i t a t i v e measures of the d o m i n a n c e of the principal trees were b a c k - p l o t t e d o n t o the t w o - w a y classification table to define c o m m u n i t i e s in terms of their c a n o p y c o m p o s i tion. A g o o d r e l a t i o n s h i p was f o u n d a n d v a r i a t i o n s in this are discussed. D i r e c t g r a d i e n t analysis of c o m m u n i t i e s was d o n e by p l o t t i n g the classified sites o n t o a l t i t u d e / a s p e c t axes. This p r o v e d useful in i n t e r p r e t i n g the g r o u p s as m a p p a b l e units. The e n v i r o n m e n t a l range of the m a i n trees was defined by p l o t t i n g their i m p o r t a n c e as relative basal area on these axes. I n d i r e c t g r a d i e n t analysis by p r i n c i p a l c o - o r d i n a t e analysis of sites and species using p r e s e n c e / a b s e n c e d a t a p r o v e d d i s a p p o i n t i n g . W h i l s t e m p h a s i z i n g the inherent c o n t i n u u m c h a r a c t e r of several classification groups, the analyses a p p e a r e d to suffer f r o m n o n - l i n e a r d i s t o r t i o n a n d did not suggest u n a m b i g u o u s relationships with any direct e n v i r o n m e n t a l gradients or resource gradients. The v e g e t a t i o n c o m m u n i t i e s of T i d b i n b i l l a N a t u r e Reserve are briefly described and related to the local climate a n d geology. C l a s s i f i c a t i o n a n d direct g r a d i e n t analysis e m e r g e d as the m o s t v a l u a b l e aids to m a p p i n g a n d the hierarchical nature of the first m a y allow systematic simplification at smaller scales.
Introduction
Background * Nomenclature follows Burbidge, N. T & Gray, M., 1970, 'Flora of the Australian Capital Territory', ANU Press, Canberra. ** Acknowledgements: This work was undertaken as part of the scientific program associated with the development of Tidbinbilla Nature Reserve near Canberra. This paper was written as a result of encouragement by the late R. H. Whittaker to whose memory it is dedicated. The assistance of Richard Deutkewitz, Paul Marchant, Terry Thomas and Kevin Hills in field work and of David Thompson, Barry Griffiths, Don Butt and Jan Ward who worked on data analysis and field checking of the map is recalled. Advice from Dr Mike Austin and Peter Milne of CSIRO was generous. Final drawings for this paper were prepared by Cangraphics and Microdata Canberra. Vegetatio 51, 157--179(1983). © Dr W. Junk Publishers. The Hague. Printed in The Netherlands.
S o u t h - e a s t A u s t r a l i a n m o u n t a i n vegetation, alt h o u g h d o m i n a t e d by a single genus, Eucalyptus, exhibits c o m p l e x p a t t e r n s of c o m m u n i t y variation. V e g e t a t i o n f o r m a t i o n s range f r o m tall closed forests to low w o o d l a n d a n d even mallee (Costin, 1954). R e g i o n a l l y , c o n s i d e r a b l e v a r i a t i o n occurs as c o m m u n i t y c o m p o s i t i o n r e s p o n d s to b r o a d climatic a n d a l t i t u d i n a l gradients. Locally, small areas such as T i d b i n b i l l a valley m a y be very diverse, as a l t i t u d inal gradients interact with m o r e a b r u p t changes in aspect, geology a n d drainage.
158 Because major centres of population have ready access to most of this region, demand for conservation-recreation land use has increased. As well as production forests (Calder & May, 1975) tracts of land ranging from wilderness to resumed farmland are now in recreation or nature conservation reserves requiring detailed planning and management. In particular, knowledge of vegetation types is central to modern fire use and control plans (Kessel & Good, 1980; Ingwersen, 1977). Regional mapping is fragmentary and may only serve specialized interests, in particular, production forestry. Its small scale is usually unsuitable for small areas requiring detailed attention. Evans (1971) and Lang (1970) examined the adjacent
Brindabella Range without mapping. Although Costin (1954), CSIRO (1969) and ANU (unpubl., 1973) have covered adjacent areas and not withstanding a history of mapping in and around Tidbinbilla (Rain, 1920; Pryor, 1938; Forestry and Timber Bureau, 1952a, b) a new interpretation of the timbered communities consistent with the management aims and educational purposes of the Reserve was required. This paper describes the numerical analysis of this vegetation and provides a resume of the composition and ecology of each mappable community. The map will be independently published in colour at 1:25 000 scale.
k
~hale )he •
TIOBINBI LLA TOPOGRAPHY
/
....M
Altitude in metres
TIDBINBILLA GEOLOGY
Fig. 1. (left) Topography of Tidbinbilla Nature Reserve and distribution of sample sites (1-81). After; Australian Survey Office (1975-78). (right) Principal geological features of the study area. After; Owen & Wyborn (1979).
159
Study area
Methods
The upper valley of the Tidbinbilla River, a minor tributary of the Murrumbidgee River is located at 148°E 35°S just W of Canberra (Fig. 1). An established nature reserve today, its valley floor was first cleared for grazing in the 1880's. Tidbinbilla Nature Reserve encloses most of the valley and also a smaller one, Oakey Ck., in all 5 500 ha providing a timbered, 'natural conservation zone' above 900 m. An educational and conservation wildlife park for nearby Canberra has been developed on the semicleared and open grassland below that elevation. Within the timber, are limited areas that were logged for brown barrel ash (Eucalyptusfastigata) and coppiced for eucalyptus oil distillation (E. robertsonii and E. dives). Rangeland grazing of uncleared forest both natural and man-made fires, may have changed some structural aspects of the forests. Their floristic composition has changed mainly by gaining exotic species, especially near cleared areas and is otherwise believed to be substantially natural.
Sampling Field sampling sites were chosen systematically to allow the correlation between altitude, aspect and vegetation to be examined and used in mapping. Eighty one sites were sampled (Fig. 1) and there was some replication to encounter possible variation along a local rainfall gradient within the valley. The extremes of drainage lines and ridges were avoided. Obvious discontinuities due to disturbances within a site were avoided but whole sites were placed in areas known to have some history of logging, coppice cutting and grazing. Each site comprised 5 circular plots with centres approximately 20 m apart. Each plot was of 7.6 m radius (182 m 2) and the whole area totalled slightly less than 0.1 ha. Spread over 100 m, always along the contour except where rock outcrops or space prevented it, the plots were likely to include most species in their vicinity. The site was characterized in terms of its geology, altitude, aspect and slope.
Geology Data The geology of the valley (Owen & Wyborn, 1979) is defined by two principal units (Fig. 1). Ordovician sediments (unit Oub) overlie granitic rock, (adamellite), of the regional batholith on the west wall of the valley (unit Smf). A quartz dyke traverses the central valley floor impeding drainage from its eastern granitic slopes and supports a 'drier' type of forest. Most of this has been cleared.
Climate Rainfall and temperatur e data for the general area are summarized in Fig. 2. Thornthwaite indices (Thornthwaite, 1948) have been plotted to place the area in climatic context. Hot dry summers, cold winters and rain in autumn and late winter are expected but summers may be humid and wet in some years. Snow may lie for up to three months above 1 000 m and is expected infrequently on the valley floor. At high elevations winds from either inland arid areas or the Snowy Mountains which create a rain shadow, probably diminish the effect of precipitation although low temperatures may tend to compensate for this.
Vegetation records Species presence was recorded within each plot. As the survey was done over several seasons, variable reliability resulted with respect to infrequent or only seasonally obvious herbaceous plants. Dubious records and species occurring in one site only were discarded. The species list is therefore not complete but is a set of 129 species which comprises most of the flora of the timbered communities. The total flora of the Reserve is ca. 300 species including old pastures (Ward & Ingwersen, 1979). The circumference (girth-at-breast-height) of all stems rooted in the plot was measured (for GBH />23 cm). Stems c o m m o n to one coppiced tree were grouped allowing stem and tree numbers to be considered independently in relation to coppicing and other past disturbances. Those data are not discussed here but the percentage relative basal area of each species (RBA) has been used in defining community units and species ranges. Descriptive notes on forest structure and site conditions were made.
160 oC
Imml 200 924m IA SL i
tmmj
~mn/~
30
I 2001
2OOI" 660m [AS L
/ 100
1000
100 /
10
800 0
0 ORRORAL VALLEY
CONGWARRA
200
30 lll6MIASLI
20
30 kin;
(b)
Inlml 2OO
0(7
I
I0
ANNUAL PRECIPITATION (1971-77) ram. DISTANCE NORTH FROMORRORAL VALLEY NUMBERSREFER TO RAINFALL STATIONS
5
1
IITI[TI
7
125
100
625m IA S L 1O0
10
100
050
I0 2 HONEYSUCKLE CK
2OO
I205m
050
075
~
]00
RIVERLEA Hnnu 200
~A S L
HONEYSUCKLE CK
HONEYSUCKLE CK
125
594m IA SL j
924m (A S L)
100
1 O0
100 "
100
~
50
075
050
7
3
PJERCE S CK FOREST H Q
BACK FLAT!
200
765m 4A S L J
ORRORAL VALLEY
°C
I m m~
200
ORRORAL VALLEY
30
576mlAS L)
100
1.O0 100 ~
2 0 5IA 7S L5J m
100 075 0.50
100 050
4
JFMAMJJASOND CANBERRA IA M O )
F MAM I IASON TIORiNBILLA VALLEY PRECIPITATION MAX
MIN
TEMP
(mm)
MEAN
MONTHLY
OC
MEAN
MONTHLY
(a)
025
FMAMJJASONO CANBERRA IA M O I
F M A MJJA S O N O CANBERRA(A M OI
THORNTHWAITE P-E RATIO/MONTH
THORNTHWAITE T-E RATIO/MONTH
(c)
(d)
161
Climate records
2 75
Measurements of rainfall and temperature (where available) from Tidbinbilla and nearby stations were obtained from the Australian Bureau of Meteorology and the Department of Housing and Construction. In Fig. 2, mean monthly values have been plotted for these stations. Stations 1-7 form a sequence from south to north across Tidbinbilla showing a general increase in precipitation with altitude. Thornthwaite's (1948) indices were plotted as they suit the limited data and allow comparison with similar calculations for other regional stations discussed by Costin (1954). Using lapse rate of 0.98 °C/100 m altitude Honeysuckle Ck temperature data were extrapolated for approximately 1 400 m altitude on the Tidbinbilla Range. Precipitation was extrapolated graphically, linearly from Orroral, Honeysuckle Ck and Back Flats values (Fig. 20.
lll6M{ASLI 250
225 4O
2 O0 2O 1 75
0
150
HONEYSUCKLE CK
125 924m ( A S L /
4°1 60
1 00
075
P-£ RATIO
2O
0
40 ORRORAL VALLEY
Analysis methods
20 576m (ASLI
so /
60
Classification of floristic presence~absence data
P-T RATIO
40
0
JFMAMJJABON
The site-species matrix of presence/absence data for each site (5 plots bulked) was clustered by an agglomerative hierarchical strategy based on Information Analysis (Williams et al., 1966). A C.S.I.R.O. TAXON Library program used the change in the Shannon information index as the measure of similarity. Hierarchical clusters were formed by combining individuals and groups of sites whose combination resulted in the least gain in information content (I) by the resultant group.
D
CANBERRA {A M O I 00
T E RATIO
THORNTHWAITE P-T R A T I O / M O N T H
{e) I = sn log n - E l[aj log aj + (n-aj) log (n-a~] j= 1O0
0
JFMAMJJASON PRECIPITATJON(mmW TEMPI°C) MEAN MONTHLY EXTRAPOLATED VALUES FOR TIDRINBILLA RANGE, PRECI PITATION, TEMP. AND THORN THWAITE RATIOS
(f)
where I is the 'information content' of the data set and there are aj individuals possessing attributej in a group of n individuals with s attributes. The resulting dendrogram can be systematically
Fig. 2. (a) Climatic data for study area and nearby stations; (b) Rainfall along N-S transect through Tidbinbilla and nearby stations (excluding Canberra); (c), (d), (e) Climatic patterns for nearby stations; (f) Extrapolated patterns for approx, l 400 m altitude on Tidbinbilla Range.
162 O0
I
[3
Fig. 3. Clustering hierarchies; vertical axis is information content increment, 2~1;(A) species, (B) sites.
truncated at any chosen information gain level to define terminal clusters. Transposing the matrix permits inverse analysis, clustering the species (Fig. 3).
on the presence/absence data using the 'Canberra metric' defined as:
J
[([Xlj xzjt)/(x~/+ x2j)]
Direct gradient analysis
Where, for a set ofn individuals having s attributes,
Direct gradient analysis (Whittaker, 1967) of the classified site groups was done by plotting each site, group-coded, with respect to the altitude/aspect axes (Fig. 5). By direct gradient analysis the ranges of the tree species were estimated using a graphical method (Fig. 6). Circles with diameter proportional to relative basal area were plotted against altitude and aspect for each site.
Xlj and x2j are the values of individuals I and 2 for thejth attribute. In this analysis, the principal diagonal was unity and x was replaced by (l-x). The Gower adjustment was performed whereby a positive matrix was produced and a principal components reduction of the datas dimenionality is effected (Williams, 1976). Results
Indirect gradient analysis (PCOA) Principal coordinate analysis was carried out on the floristic data matrices (inverse and normal) using the program GOWER (C.S.I.R.O. Division of Computing Research, TAXON Library). In this procedure, a similarity matrix was computed based
Classification of sites and species on presence~absence data The dendrogram of site clusters (Fig. 3 B) revealed ll interpretable site groups, nine definable at a
163 consistent information level and four as subgroups below this. As there is no statistical criterion for this truncation, ecological interpretability was used to the limit division. For species (Fig. 3A) 10 groups at one level, with two lower sub-divisions, provided clusters that associated recognizably with site groups when the data were replotted in the order now provided by this analysis (Fig. 4). A summarised description of the communities defined by the table and other data is provided in a later section of this paper. Truncation of each array of clusters depended on some experience of the vegetation concerned. Species groups were sought which appeared to have the most clearly defined correlations with particular site groups. Within species groups (Fig. 4) the dominant trees of the study area were marked (t) and the principal shrubs (s) to assist in forming a summarized floristic definition of each community. It is possible for several species groups to correlate sufficiently with a site group to define a community. Where some idea of the environmental conditions pertaining to that site group can be detected, there is a good basis for defining a broad community/ habitat relationship. The vegetation descriptions are based on these principles. The broad environmental characteristics of the site groups are outlined below. All sites in groups A, B, C, D are below 1 300 m, low and usually on granite which forms most of the lower slopes. Some soils are colluvial, derived from sedimentary rock as well as underlying granite. All of group D except site 74 are on granite. Group A sites (Fig. 5) are on the lower slopes of the most shaded slopes and gullies usually with SE aspects. A characteristic 'wet gully' shrub association occurs under Eucalyptusfastigata. Group B sites occur in a similar altitude/aspect range but lack the extreme shelter from isolation and dry westerly winds provided for group A sites by the Tidbinbilla Range. They also occur further from the more mountainous end of the valley which is moister (Figs. 1, 2b). Groups C and D represent the lower footslope and valley floor environment, group D sites being largely restricted to the warmer and drier side of the valley. The high horizon of the Tidbinbilla range has less effect on these sites than those of group C. Group D sites extend higher than those of group C possibly due to the fact that the eastern slopes are
granite throughout. Group C sites may be slightly constrained by the geological boundary on the western slopes but competition from other moist site communities could be as important. Groups E/E~ and F / F l contain sites mainly at the drier end of the valley or on exposed dry aspects on sediments at the moister end. They are dry sites. Group E / E l sites are all on sediments except site 47 and group F / F l sites are on either geological substrate. Group G contains the highest sites and sedimentary rocks may secondarily act to influence the definition of this group. There are also high altitude sites in groups H and l on granite on southern aspects with different dominants. Eucalyptus delegatensis is absent from group G, strongly implying that it is favoured by granitic rather than sedimentary-metamorphic substrate. It occurs on nearby ranges within this altitude zone. Group G species are only weakly differentiated with respect to site type. Sites 22, 27, 55 and 56 (Fig. 4) would split at the 10 group level. These show some floristic differences from the other members. They are on N-NE aspects and are associated with species groups 2b, 3 and 4 but not 10b. These sites are on the northern Tidbinbilla Range, probably tending towards dryness. Groups H and I fall within an altitude/aspect range associated with the cooler, moister but not extremely cold microclimate on the mid-slopes of the Tidbinbilla range. They occur on both geological types but only site 39 is in the N W - S W range. Group I occurs on NE-S aspects and only extends northwards where a SE aspect maintains conditions equivalent to those in the central valley (sites 54 and 80). Group H occurs on drier, NE-NW sites and one (44) is high on the eastern side of the valley with a NW aspect.
Direct gradient analysis of tree stem data The results of the stem size and composition analysis were used in two ways. In Fig. 6, RBA has been plotted by site group to reveal strong quantitative relationships between tree species composition and the site groups which had been independently defined by only presence/absence floristie data. This largely confirms the site clusters as meaningful community units. As the presence/absence agglomerative classification uses all
164
A
Groups
C
B
D
Site Numbers
1
2a
2b
Asperula scoparia Poranthera microphylla Stackhousia monogyna Viola betonicifolia Poa spp. Stellaria pungens Senecio quadridentatus Acacia dealbata Clematis aristata Geranium sp. Hypochoeris radicata Luzula campestris Eucalyptus bridgesiana Acrotriche serrulata Hypericum gramineum Pimelea linifolia Plantago varia Centaurium erythraea Cksium vulgate Glyciue clandestina GnaphalJum gymnocephalum Oxalis corniculata Eucalyptus viminalis Eucalyptus rober tsonii Hydrocotyle sp. Hovea heterophylla Hardenbergia violacea Pultenaea procumbens Tetratheca bauerifolia Daviesia mirnosoides Monotoca scoparia Stylidium graminifolium Acacia falciformis Cassinia aculeata Eucalyptus dives Dianella revoluta Hibbertia obtusifolia Haloragis sp. Eucalyptus nortonii Eucalyptus rubida Acacia melanoxylon Ajuga australia Anagallis arvensis Agropyron scabrum Crassula sp. Cheilanthes tenuifolium Clematis microphylla Danthonia sp. Dodonaea sp. Dipodium punctatum Exocarpos cupressiformis Lepidosperma sp. Myosotis sp. Eucalyptus fossil Eucalyptus mannifera Eucalyptus macrorhyncha Acacia gunnii Acacia rubida Bossiaea foliosa Brachycome rigidula Billar diera scandens Choretrum pauciflorurn Cynoglossum australe Cassytha sp.
, S
S S t
S
5
S
S
S
*
* * *
s S s S
S S s s I t
t
t •
*
*
*
*
*
*
*
*
t *
* *
t
t
S
t t t t t ~ t t
t
* *
S
S
S
S
S
S
S
t~
* *
*
t t t t s
s s
S
S
S
S
S
S S
S
S
Fig. 4. T w o - w a y c l a s s i f i c a t i o n o f species a n d sites; p o t e n t i a l l y d o m i n a n t trees (t), p o t e n t i a l l y i m p o r t a n t s h r u b s (s) a n d all o t h e r s (*).
~
-
V~
g
Z
c~
~rr~
A
Groups
166
Site Numbers
B
C
5 13 I5 24 26 34 29 31 49 61 62 77 79 2 5 3 5 3 7 4 2
6
D
1632
I
4
9 17 38 40 50 48 63 (-4 2 28 36 70 71 74
Danthonia pailida Exocaxpos strictus
S
Grevillea lanigera lndigofera australia s
Leptospermum juniperinum
s
Leucopogon virgatus Melichrus urceolatus Rosa rubiginosa Solenogyne gunnii Lolielia gibbosa * * * *
Themeda australia
*
Xanthorrhoea australia Eucalyptus fastigata
t
Bedfordia arborescens
S S S S S S
t t t S
S
S
s
Coprosma quadrifida
6a
Olearia argophyUa
Ss
SS
Pomaderris aspera
S
S
Polystichum proliferum
* * * * * * * *
Is
S
s
S
S
s
Ur tica incisa Amyema pendula * * * * *
Chiloglot tis sp.
6b
Acianthus sp. Persoonia chamaepeuce Pterostylis longifolia
7
Lagenifera stipitata
* * * * *
Olearia lirata
s S S S S
Pteridium esculentum
S * * * * * * *
* * * * * *
* * * *
**
s S S S
ss
S
* * * * * * * * * *
Veronica calycina
*
Echinopogon cheelii
* * * *
*
*
Acaena anserinifolia
* * * *
*
***
S
S
• I
•
S
S
*[*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
S
* * * * * * * * *
*
Cymbonotus lawsonianus Dichondra repens
*
Epilobium sp. *
Orreomyrrhis sp. Bursaria spinosa
SS
S S
Coprosma hir tella
S S
S
S
S
s
SSS
s
s
* * * * *
Rubus triphyllus
*
*
*
**
*
***
Senecio linearifolius Viola hederacea
* * * * *
*
*
*
*
*
*
**
Acacia obliquinervia Brachycome scapiger Bulbine bulbosa s
Daviesia ulicifolia Goodenia hcderacea Oxylobium ellipticum Oxylobium alpestre Parahebe derwentiana Picris hieracioides Scleranthus biflorus Polyscias sambucifollus Trifolium repens Trifolium arvense Eucalyptus pauciflora
lOa
Craspedia glauca Euphrasia sp. Parahebe perfoliata I-ucalyptus dalrympleana
t t
t
t
t
Brachycome aculeata * * * * ~
Dianella tasmanica
10b
Helichrysum scorpoiides Lomatia myricoides Olearia erubescens Olearia megalophylla
s
s
s
* S
s
S
Galium ciliare Asplenium flabellifolium
*
S
s
*
*
* S
S
s
s
El
E 14 ~0 33 57
73
F
G
FI 3
52 59 65
H
76 22 27 55 56 51 6768
23 46
58 66 69
39
167
I
[0 1 2 1 8 4 1 4 4 4 5
s
8
II
1943
t
t
t
7
21121
54
7"~ u "
s
S
,
,
*]
S S S S
S
S S
S
*
*
S
S
S
s
S
S
*
*
*
S
s
t t t
t
s
s
t
t
S
t
•
S
t
•
S
l
•
t
t
S
S
S
S
S
S
t
t
t
1
t
•
,
S
5
S
S
S S
5
s
t
S
S
S S
S
S
S
t
t
t
t
l
t
S
t
S
t
t
1
t t t t
I
•
s s s s
S S s
S
s
168
1500
G 67
G 69
G 66
1400
G
( ''H~'}
2/
55
51
1200
}--
r~ <~
:;
=
\(~o)-
~)
-'"~,"o
=o
=
1000
6(~
F1
B
59
/
//
R
/ F~
/
42
/~, ~
\
~T
c
__
24
A
~
r-,\
74
\ 75
c
c
~
c
800
F
I 180
/
I
270
D 2 C
\
360/0
,
16,
90
180
A S P E C T (DEG. M A G . )
Fig. 5. Direct gradient analysis of sites coded A-1 by groups as in Figs. 3B and 4; numbers as in Fig. 1. species to define groups, the characteristic c a n o p y species have been chosen to name the communities. In a few cases, the species may be absent f r o m particular stands t h o u g h the characteristic shrub understorey and g r o u n d cover are present e . g . E . fastigata, absent from two stands of c o m m u n i t y group A (Fig. 4). In Fig. 6, the quantitative response of trees to microclimate expressed in terms of altitude and aspect has been plotted. E. bridgesiana is never I00% d o m i n a n t and is centred on NE aspects below 1 100 m. E. dives is the most tolerant species centred on N E - N W aspects between 1 000 m and 1 350 m but spreading onto E-SE and S aspects at lower elevations often with
high dominance. E. robertsonii is restricted to NES aspects below I 100 m and gaps in its distribution are occupied by E. dives. As these two species have been selectively cut-over for eucalyptus oil, considerable variance is associated with the measures f r o m some sites and they may not correlate as expected with the principal trends. The dominance of E. robertsonii is uneven within its range. E. viminalis occurs up to 1 200 m on N W aspects but is limited at lower elevations to SE-S aspects. Nearer the centre of its range, 750-900 m, it may be completely d o m i n a n t on SE-S aspects. E. fastigata attains almost complete d o m i n a n c e over its restricted range of 900-1 200 m on E-S aspects. Widespread between 900 and 1 400 m, E. dalrympleana becomes
169 1500
1500
E. viminalis
E. fastigata 1400
1400
1200
O
1200
O
O o
O 1000
O o
o® O
8OO
800 i
I
1500
1500
E. dalrympleana
E. robertsonii
O
O •.•
1200
o
O
1200
o
o
o
o
g
~)o
0 0
o o
O
o O
looo
,ooo O
,
o
O
0Ooo 80O
804)
=
i
1500
E, dives
15~}
14100 1400
o 1200
O
O0 0 E.pauciflora 0 O Q O
1000
0
0
o%°° %0o
o 80~
i
i
150~ E
E. bridgesiana
1400
W
O
1200
O
1000
O %
80O S
O
o
o
o
100
75
percent
"
O
5o
"
o
25
"'
R e l a t i v e basal a r e a f o r seven d o m i n a n t Oo
S
ASPECT
t r e e species
TIDBINBILLA NATURE RESERVE
ASPECT
Fig, 6. Direct gradient analysis at performance of principal dominant trees.
170 SITE GROUP SITE NUMBER SPECIES
A
B
C
D
E
El
F
F,
G
H
I
E. bridgesiana E. dives E. robertson#
j
- -
E. viminalis E fastigata E dalrympleana £ pauciflora
,,_Lee
-,,
i• 1 1 1
.
.
_
.
I
.
,
,_..,
._
L
d. ~ - - ! ~ i
E. rossii E mannifera
L,
I
.11
m
=11
II!
E. macrorhyncha A falciformis I
--
m E
--
I
E cupressiformis
E delegatensis
Relative basal area of principal dominant trees arranged by sitegroups Fig. 7. Relative basal area of principal dominant trees arranged by site groups.
more d o m i n a n t at the higher elevations and on cool E-S aspects at the lower levels. Here, it occupies moister sites below E. pauciflora which are more sheltered and where the catchment above provides g r o u n d w a t e r run-on. E. pauciflora occurs on all aspects above 1 200 m where it is almost always highly dominant. Differences in stand composition are small but there are differences in m o r p h o l o g y between trees and in stand structure on contrasting aspects within this single dominant.
i
,,'i
I_
E nmtoni E rubida
_.J
Indirect gradient analysis by principal coordinate analysis (PCOA) The sites are displayed against the P C O A axes in Fig. 8. Boundaries defining their classification groups are also included. To interpret the axes, some information derived f r o m the classification was also used. Only two axes seemed to be interpretable. G r o u p G, associated with coldness at high elevations and probably moisture stress due to exposure contrasts on axis I with groups A and B. These are on sheltered, w a r m e r slopes and g r o u p A is restricted to moist sites.
>
171 mesic
4
s~u
Rs • B's
lc.
MaAA'm
F~ 2 pol
4o sq
warm
-4
--2
II ~oid
B°r
-.2
Ro
Fig. 8. Indirect gradient analysis (PCOA) of sites; axes I and II.
A ,.6
xeric
Groups H and I are cool sites at intermediate elevations. Group H, on warmer, drier aspects of ridges alternates with group I within the same altitude range. Groups E and F are in drier areas than C and D or H and I and occur in the positive region of axis II. Groups E, F and D are warmer sites than either A, B and C or the upper elevation groups G, H and I. The axes may superficially be equivalent to altitude (I) and aspect (II). Given the complex correlations between altitude and rainfall, northern aspects and higher temperatures, axis I may tentatively be characterized as a moisture factor gradient. Axis II may similarly be related to temperature. The majority of the groups lie on a 'horseshoe' sequence from lower left to right then upper right to left. Non linear distortion resulting in an 'arch' pattern in points analysed by P C O A has been discussed by Austin (1976a, b). This possibility lessens the value of this analysis in describing the vegetation. Inverse (species) P C O A is difficult to interpret. Twenty-six species have been selected for display in Fig. 9 as something is known of their ecological tolerances.
Fig. 9. Indirect gradient analysis (PCOA) of species selected for their 'indicator value'; Af, Acaciafalciformis; Ad, A. dealbata; Am, A. melanoxylon; Ar, A. rubida; Bs, Bursaria spinosa; Br, Eucalyptus bridgesiana; Dp, Danthonia pallida; Du, Daviesia ulicifolia; Dm, D. mimosoides; Da, E. dalrympleana; Dr, Dianella revoluta; Dt, Dianella tasmanica; D, E. dives; F, E. fastigata; Ma, E. mannifera; M, E. macrorhyncha; P, E. pauciflora; Pp, Parahebe perfoliata; Pd, P. derwentiana; Pol. Polystichum proliferum; Ro, E. robertsonii," Rs, E. rossii; Ru, E. rubida; Scb, Scleranthus biflorus; Sp, Stellaria pungens; Sq, Senecio quadridentatus.
The acacias form a series with respect to axis I. Acacia rubida and A. melanoxylon towards its positive extreme contrast with A. dealbata at the other. A. rubida is widespread in the colder mountain areas and A. melanoxylon occurs mainly in cool areas but usually thrives only where there is readily available ground water-drainage lines and flush zones around granite tors. A. dealbata occurs widely at lower elevations, mainly on moister footslopes whereas A. falciformis is found on warmer drier slopes. Eucalyptus fastigata and E. pauciflora lie towards the positive end of axis II and E. dives, E. robertsonii and E. bridgesiana occur towards the negative end of each axis. Within this group, E.
172
bridgesiana is remote from the other two, near the negative extreme of axis II. E. robertsonii and E. dives are cool-warmer footslope species, the latter being more tolerant of dry sites and very widespread. Similarly, E. bridgesiana occurs in even drier and warmer areas than E. dives. The trends within Acacia and Eucalyptus imply that axis I is related to moisture and axis II, temperature. On the other hand, E. rossii, E. macrorhyncha and E. mannifera do not fit this pattern. This may be on account of the small number of sites in relation to the number of species involved. Other species, pairs within the same genera, reinforce the interpretation suggested. Dianella tasmanica and D. revoluta occur in relatively cool, moist and warmer, drier habitats, resp. Parahebe perfoliata is found in colder drier areas (snowgum community) than P. derwentiana. Daviesia mimosoides is extremely widespread in dry, warm areas and cold dry sites also whereas D. ulicifolia is confined to cold areas and probably requires moister conditions. Eucalyptus dalrympleana and E. viminalis occur at high and low elevations resp. suggesting greater cold tolerance in E. dalrympleana. They are widely separated along axis II (temperature) but not much on axis I (moisture factor).
It is largely distinguished from related communities by its understorey and Eucalyptus fastigata which may occur with E. dalrympleana in its upper range and E. robertsonii in adjacent lower elevation areas. The characteristic tall shrub understorey contains species typical of the wettest cool eucalypt forests of south eastern Australia; Bedfordia arborescens, Coprosma quadrifida, Oleariaargophylla, Pomaderris aspera, Polystichum proliferum. Other associates of this community occur in species groups l, 2b, 6a, 7 and 8b (Fig. 4). Infreqent species in this community which were either not sampled or discarded (because they failed to occur more than one site) included Culcita dubia and Dicksonia antarctica (abundant along some creeks but not sampled) and Polyscias gaudichaudiana. Structurally, the community has a 20-30 m high canopy over an irregular shrub layer up to 5 m high with relatively bare ground occupied by Pteridium esculentum especially in gaps. The herbaceous ground flora is diverse but sparse.
Eucalyptus robertsonii community Site group: B Dominant trees: E. robertsoniL E. viminalis, E. dives, E. bridgesiana Principal shrubs: A.caciafalciformis Major ground cover: Poa spp, (tussock), Pteridium esculentum
Ecological description of Tidbinbilla vegetation
Eucalyptus fastigata community Site group: A Dominant trees: E. fastigata, E. viminalis, E. robertsonii, E. dalrympleana Principal shrubs: Bedfordia arborescens, Olearia argophylla, Pomaderris aspera Major ground cover: Polystichum proliferum, Pteridium esculentum
This community occurs on slopes similar to those of the E.fastigata community (group A). Sites have a more northerly aspect and occur further towards the drier northern central valley, generally between 820 and 1 070 m in altitude. It is closely related to the wetter forest (Fig. 3B) and it shares E. robertsonil and E. viminalis and many understorey species but not the distinctive group 6a 'wet sclerophyll' shrubs. On drier slopes, E. dives and E. bridgesiana may be significant components. Such sites indicate gradation into the related and often adjacent drier footslope communities (groups C and D, Figs. 5 and 8).
This community which favours low radiation sites is sheltered from westerly winds in the moister part of the valley and lies generally between 820 and 1 070 m in gullies and on the eastern footslopes of the Tidbinbilla Range.
Eucalyptus robertsonii - E. dives E. viminalis community Site group: C Dominant trees: E. robertsonii, E. dives, E. viminalis
Communities of the lower slopes and valley floor
173
Acacia dealbata, A. falciformis, Olearia lirata Major ground cover: Pteridium esculentum
Principal shrubs:
In this community between 760 and 1 070 m altitude on the western and southern footslopes, E. robertsonii is more dominant than E. dives. E. viminalis, dominant along drainage lines, is often present. The community is fairly sheltered but receiving more radiation than the preceding types, it may be regarded as warmer and drier. It falls between B and D in the P C O A ordination (Fig. 8). The main species are in groups 1, 2a, 4, 7, 8a and 8b. I m p o r t a n t shrubs are Cassinia aculeata and Bursaria spinosa. Ground cover is diverse with the ericoid Acrotriche serrulata, Hypericum gramineurn, Pimelea linifolia, Centaurium erythraea, Dianella revoluta, Hibbertia obtusifolia, Poa spp. and Pteridium esculentum. The last two vary in abundance within and between sites and this is not revealed by the presence/absence data analysed here. 'Poa" may include up to three species.
Eucalyptus bridgesiana - E. dives community Site group: D Dominant trees: E. bridgesiana, E. dives, E. rubida Principal shrubs: Bursaria spinosa Major ground cover: Poa spp. This community contains smaller amounts of E. robertsonii and E. viminalis which suggests a drier type of vegetation. E. rubida is occasionally present, a species of limited occurrence at Tidbinbilla. The main difference between this community and group C is quantitative in the tree stratum. Most shrubs and herbs are c o m m o n to each with mainly quantitative differences not shown by these analyses. Themeda australis, Acaena anserinifolia and Cymbonotus lawsonianus are group D species. With a higher upper limit, this community's altitude range encompasses that of group C. Its lower limit was found, on field checking, to rise towards the southern moister, sheltered valley area as group C replaces it. Only site 74 was not on granite and the eastern slopes are generally not protected from westerly wind. Solar radiation is high and the vegetation shows fairly uniformly, the characteristics of a dry habitat.
Groups C and D then, are the warmer slope and valley floor communities both on granite, differing mainly in the mix of dominants. They occur on all but N-NW aspects within a similar altitude range and generally on slopes of 10-25 degrees.
Communities of upper northern valley slopes Eucalyptus dives community Group: E Dominant trees: E. dives Principal shrubs: Daviesia mimosoides Major ground cover: Poa spp. On sedimentary rocks between 920 and 1 220 m where there is m a x i m u m radiation and least shelter, this community is often pure E. dives. The main shrubs are Daviesia mimosoides, D. ulicifolia, Acacia rubida and A. falciformis. The ground cover of low heathy shrubs includes Hovea heterophylla, Tetratheca ericifolia, Monotoca scoparia, Bossiaea foliosa and Hibbertia obtusifolia. Tussocks of Poa spp. and some of Danthoniapallida form an open graminoid community enriched by Stylidium graminifolium, Daniella revoluta and less c o m m o n herbs. Understorey varies in composition and density.
Eucalyptus rossii - E. macrorhyncha - E. mannifera community Group: El Dominant trees: E. rossiL E. macrorhyncha, E. mannifera, E. dives Principal shrubs: Acacia rubida, Daviesia mimosoides Major ground cover: Danthonia pallida Although not extensively sampled, this locally restricted community is widespread in the region on the hills of the dry lowland plains, especially near Canberra where it has been well described ( P o o k & Moore, 1966). This group is centred on more northerly aspects than group E which has a westerly bias. At Tidbinbilla it characterizes the driest low forests of the main range and is geographically the closest community to the warm, dry, tablelands. Altitude range is as for group D. Many of its shrubs are c o m m o n to group E. Of special interest is the arborescent graminoid, Xanthorrhoea australis, which often persists in grass-
174 land formed where this and related communities have been cleared (Gill & lngwersen, 1976). Indigofera australis, Acacia rubida and Cassinia aculeata usually occur.
Eucalyptus dives - E. mannifera community Group: F Dominant trees: E. dives, E. mannifera, E. robertsonii Principal shrubs: Aeacia falciformis, Cassinia aculeata Major ground cover: Danthonia pallida On the northern Tidbinbilla Range, are two sites on sediments and on aspects which carry E. dives and various associated trees. Floristically similar to the footslope groups C and D, they are drier, receiving more radiation. Soils are stony, skeletal or deeper, coarse textured. E. mannifera does not occur in groups C and D. The understorey is shrubby. A. falciformis and Cassinia aculeata occur irregularly with a tussock cover of Danthonia pallida. Themeda australis is sometimes present and on land cleared from forest of this type to develop pasture, it has become one of the main dominants.
Eucalyptus dives - E. bridgesiana - E. nortonii community Group: F~ Dominant trees: E. dives, E. bridgesiana, E. nortonii Principal shrubs: Xanthorrhoea australis Major ground cover: Danthoniapallida Small areas of dry forest similar to the preceding types occur on N-NW, warm, dry aspects at the northern end of the valley. Of five sites in this group, three are on granite, one is transitional and the remaining one is on sediments. On granite, on the eastern slopes, E. nortonii distinguishes this community, occurring in various proportions with its co-dominants. All trees are of poor form. The habitat is typically dry on coarse granitic soil. Within the same altitude range as E. bridgesiana, 760 to 1 070 m, E. nortonii occupies more arid, warmer sites on granite.
Communities of upper central and southern valley
slopes Eucalyptus dalrympleana Group: H
E. dives community
E. dalrympleana - E. pauciflora community Group: I As they largely occupy contrasting aspects of ridges on the Tidbinbilla range between 1 100 and 1 300 m, these communities will be discussed jointly. On NE and N-SW aspects, group H is dominant and on cooler NE-S aspects, group I prevails. Dominant trees: E. dalrympleana, E. dives,
E. fastigata, E. pauciJTora Principal shrubs: Drier (group H)
Moister (group I) Lomatia myricoides Olearia erubescens Coprosma hirtella Olearia megalophylla Major ground cover: Pteridium esculentum, Poa spp. Other species showing some habitat preference: Glycine clandestina Senecio linearifolius Oxalis corniculata Viola hederacea Hardenbergia violaceaBrachycome aculeata Veronica calycina Bark, twig and leaf litter may be the main ground cover especially under the purer stands of E. dalrympleana. The drier community usually abuts the upper limits of footslope communities with E. dives. Group H sites are within this tree's range (Fig. 7). The moister community abuts E. fastigata or E. robertsonii and E. viminalis stands below and so may contain significant amounts of these. Altitudinal boundaries are less clear than those formed by topographic changes in aspect. Above each community, E. pauciflora is the single dominant of the sub-alpine zone.
Daviesia ulicifolia Cassinia aculeata
Communities of sub-alpine slopes and ridges Eucalyptus pauciflora community Group: G Dominant trees: E. pauciflora
175 Principal shrubs:
Oxylobium ellipticum, Olearia erubescens, O. megalophylla, Acacia dealbata, Daviesia mimosoides and Coprosma hirtella. Also significant but less c o m m o n and not generally sampled by this survey are Eriostemon lanceolatus, Drimys spp. Oxylobium
alpestre, Leptospermum myrtifolium and Leucopogon suaveolens. Major ground cover: Poa spp., herbs and low shrubs. Above 1 220 m, E. pauciflora is the single dominant tree. No other available species seems able to compete with it or withstand the coldest extremes of the environmental conditions found in this region. Dryness rather than moisture is coupled with this to increase environmental stress and soils are skeletal, acidic and receive less ground water runon by comparison with group H and I sites below them. Rock outcrops are c o m m o n and support E. pauciflora with roots penetrating crevices. Contrasting aspects affect tree morphology and density rather than community composition. Trees are dwarfed and twisted on dry windy slopes but denser and taller on sheltered ones. Further south in the main mountain block of the ACT and nearby Snowy Mountains, E. pauciflora associates with E. rubida over broad areas at lower elevations but this association is u n c o m m o n at Tidbinbilla. E. rubida occurs at high elevations on sediments on the northern part of the Tidbinbilla Range. At low elevations it occurs with several species of eucalypt along the Tidbinbilla River and adjacent footslopes. On dry slopes, E. pauciflora commences higher at 1 220 m, abutting the E. dives community. On moist cold slopes its lower limit is 1100 m where it passes into group H or I sites with E. dalrympleana. Floristically diverse ground cover species call attention to this community as well as its attractive variable tree morphology and regular snow cover. It is central to nature conservation and sub-alpine recreational interests in the region. Amongst these species are Bulbine bulbosa, Da-
viesia ulicifolia, Parahebe derwentiana, P. perfoliata, Stylidium graminifolium, Craspedia glauca,
Euphrasia spp., Dianella tasmanica, Brachycome spp. and Helichrysum scorpioides. Species groups 1,8 (occurring in other communities) 9, 10a and 10b (more typical of group G only) contain the characteristic flora.
Eucalyptus delegatensis community Group: Dominant trees: Principal shrubs:
I
E. delegatensis Acacia dealbata and as in
groups 8, 10b Major ground cover: Acacia dealbata and as in groups 8, 10b. This tall forest occurs on very few sites, mainly peaks and ridges above 1 190 m on granite only. At this altitude at Tidbinbilla the only granite available is on unfavourable aspects and occurrence of the tree is limited by geology within the study area. It is a widespread sub-alpine saw-log tree of south eastern Australian mountains.
Discussion
Biophysical relationships The aim of the study was to generate vegetation map units in a way that reflects major environmental relationships. Real data on microclimatic gradients, ground water or soil features were not available so geographical parameters related to them and capable of being used to extrapolate between sites on a topographic map have been emphasized. These are 'indirect environmental gradients' (Austin & Cunningham, 1981). Classification analysis showed that although the clarity of the relationship was not the same for all units, identifiable land categories were associated with particular vegetation units. The ordination by P C O A showed that this lack of clarity was attributable in part to the continuous variation encountered in vegetation along altitude and aspect gradients. The apparent sharpness in vegetation change on traversing ridge and gully sequences is reduced by the analysis as the points are displayed out of geographical sequence in abstract. As several primary factors or 'resource gradients'
176 (Austin & Cunningham, 1981) may be confounded, biophysical relationships can only be suggested by a study such as this. The geographical correlations can be stated more directly although their explanation may be incomplete. Biophysically, increasing coldness at higher elevations is important in sorting species into major community zones. On ridges and peaks, groundwater must be reduced by run-off. Precipitation, although higher, is less effective due to wind. In summer, storm rains and, in some years, general rain and cloud lessen the seasonal stress that must occur even at high altitude in mountains in this region. This accords with the description of the snowgum woodland of the Shoalhaven area to the east as 'dry' (CSIRO, 1969). In addition to the local gradients reflected in altitude and aspect changes, there is a rainfall gradient from south to north along the valley. The available data (Fig. 2b) show that precipitation drops with altitude and distance from the southern area of continuous mountain land. Station 3, Back Flats represent this area. The sharp drop at number 5 could reflect a rainshadow caused by Mt Tidbinbilla. Stations 6 and 7 are close to the extreme northern ridges, which being lower than the summit do not impede rainfall as much. The sampling design included sites with equivalent altitude and aspect throughout the length of the valley to encounter variation such as this. Group A tends to occupy more southern sites on the western slopes than group B on similar aspects. Group F occupies eastern slopes and on the western side, some more northerly sites than group E which is also widespread. Group D occupies the eastern footslopes and group C the western ones but on the latter, the more northern equivalent sites are in group D. These trends maybe related in part to increasing dryness towards the north.
Community nomenclature and map units Classification has been emphasized because of mapping. The choice of geographical correlates such as aspect and altitude as well as a knowledge of geological variation allows classified points to be extrapolated on topographic map, a parametric mapping approach (Mabbutt, 1968). When supported by analyses of the continuum character inherent in this vegetation, the unavoidable simplifi-
cations that boundaries must cause should not greatly detract from its value. Interpretation was checked extensively on the ground and it was more difficult to define a community name than to recognize fairly clear changes in vegetation along traverses. To generate a useable nomenclature, the eucalypts contributing consistently, or if less consistently, with high RBA (Fig. 7) were selected. These names are approximately equivalent in status to the 'associations' of Pryor & Moore (1955) and Costin (1954). They are tags with some descriptive value, no more than that. Important associated species are listed in the community descriptions and in these, the more detailed composition of the stands is outlined. To map the units, the range of each site group was defined with reference to Fig. 5 and land falling within these limits was coded accordingly. At this stage, some adjustments were made on the basis that high altitude units extend lower on cold slopes and drier units extend higher on warm aspects. The geological dichotomy was taken into account also at this stage where necessary. Then a number of field traverses were made to check the primary map. Allowing for some weakness of definition and continuum aspects, the vegetation generally corresponded with the map and its principles. Further adjustments were made. The mapped boundaries of the E.fastigata community were adjusted and the limited areas of E. delegatensis were mapped from air photographs. Photo interpretation was not used to map elsewhere except in checking and defining some boundaries. Uniform interpretation at the level of resolution required would have been difficult on account of heavy shadowing in the steeper terrain. Structural rather than floristic differences tend to show and the two do not always correspond, hence the use of the parametric approach. The draft map at 1:12 000 scale is being prepared for publication independently at 1:25 000 scale. In the Appendix, the equivalent nomenclature of other workers is given for related areas including Tidbinbilla. In general, this study has increased detailed knowledge of Tidbinbilla Nature Reserve. Existing units (Pryor, 1938; Pryor & Moore, 1955) could not be used to sub-divide this small area with the resolution now required.
177
Evaluation of methodology The number of samples was large for such an area and careful stratification could have reduced the number. Although a measure of frequency based on the five plots per site was available it was discarded as five is too small a value for reliability. Multiple plots take more effort to lay out than large single ones and experience suggests that 0.1 ha (as 20 X 50 m) provides a 'good sample' in this vegetation. The classification program used agglomerative sorting. Divisive sorting has been tried and found inferior in producing useable association tables although there was broad agreement. The hierarchical nature of classification has advantages where map units are sought. Systematic reduction of map units in related sets is possible. For example, at the 9 group level, group E l and E combine without loss of significant information if a simpler map or analysis is required. At the 7 group level, E. fastigata is absorbed into a complex but closely related group of associations with E. robertsonii and E. viminalis. The first two splits in the hierarchy separate footslope types from higher elevation sites and drier exposed sites from high sheltered cold ones. At the 5 group level, a useful map could have been prepared suitable for maps at 1:100 000 or 1:50 000 scale. The combination of classification analysis with RBA data allowed the quantitative variation in the tree stratum to be used in checking the floristic classification of sites and in naming them (although some weaknesses exist). Direct gradient analysis (Whittaker, 1967) was applied to the classification results (Fig. 5) and this provided a basis for mapping and broad biophysical correlations. Principal coordinate analysis being an indirect gradient analysis is not readily interpreted and is susceptible to non-linear distortion. It showed the close relationship between the apparently sharply defined classification units. It was the least useful strategy of the three (classification, DGA, PCOA) in developing mappable community concepts but served to emphasize the continuum within which the definition of some units was to some extent forced. On the basis of this study, similar work could be improved by stratifying the samples with respect to the basic geological dichotomy. This was not done initially as the boundaries had to be found
in the process of sampling. Transitional soils marked it in many areas. Also, fewer samples could have produced a similar result. Although highly distinctive units like the E. fastigata and E. rossii communities require few samples for relative definition, stratification to reduce sampling in otherwise recognizable areas may create imbalance in the data matrix. The program used is apt to lump together, odd individuals that have in common their inability to classify with 'strong' groups and is sensitive to 'double zero' matches. In this vegetation and geological system, local irregularities can be considerable because of granite tors, bare rock faces and springs. Acceptance of structural variability and tree composition within a broadly recognizable community is necessary. The nature of the Eucalyptus dalrympleana communities is not clear from the study of this area and a broader study would be useful in clarifying associated dominants and habitat range.
References A.N.U., 1973. A resource and management survey of the Cotter River Catchment, Resource and Environment Consultant Group, Dep. of Forestry, Austral. Nat. Univ., Unpubl., Canberra. Austin, M. P., 1976a. On non-linear species response models in ordination. Vegetatio 33: 33-41. Austin, M.P., 1976b. Performance of four ordination techniques assuming three different non-linear species response models. Vegetatio 33: 43-49. Austin, M. D. & Cunningham, R. B., 1981. Observational analysis of evironmental gradients. In: Littlejohn, M.J. & Ladiges, Pauline Y. (eds.) Ecological gradients and boundaries, Proc. Ecol. Soc. of Australia, II, Melbourne. Australian Survey Office, 1975 78. ACT 1: 10 000 Planning Series, Sheets 184 (576-588) and 192 (576-588), Dept. of Administrative Serv., Canberra. Calder, S. W. & May, J. R., 1975. Landscape as a resource. In: Kikkawa, J. & Nix, H. A. (eds.) Managing Terrestrial Ecosystems, Proc. Ecol. Soc. of Australia 9: 250-255. Clifford, H. T. & Williams, W. T., 1976. Similarity measures. In: Williams, W. T. (ed.), Pattern analysis in agricultural science, CSIRO, Melbourne. Costin, A. B., 1954. A study of the ecosystems of the Monaro Region of New South Wales. N.S.W. Government Printer, Sydney. CSIRO Division of Computing Research - TAXON Library. 306 Carmody Rd., St Lucia, Qld, Australia, 4067. CSIRO, 1969. Lands of the Queanbeyan-Shoalhaven Area, A.C.T. and N.S.W. Land Research Ser. 24. CSIRO, Melbourne.
178 Evans, O. W., 1971. Altitudinal variation in vegetation on the Brindabella Range, B Sc. Hons. Thesis, unpubl. Dept. of Bot., Austral. Nat. Univ., Canberra. Forestry and Timber Bureau, 1952a. Unpubl. map, ACT Gibraltar Creek area, forest type map, 20 ch:l inch, dyeline, Canberra. Forestry and Timber Bureau, 1952b. As above, Tidbinbilla River area. Gill, A. M. & Ingwersen, E., 1976. Growth of Xanthorrhoea australis R. Br. in relation to fire. J. Appl. Ecol. 13:195 203. Ingwersen, F., 1977. Vegetation development after fire in the Jervis Bay Territory. MSc. Thesis, A.N.U. Canberra. Kessell, S. R. & Good, R. B., 1980. PREPLAN (Pristine Environment Planning Language and Simulator) user's guide for Kosciusko National Park. National Parks and Wildlife Service of N.S.W. Sydney. Lang, G., 1970. Die Vegetation der Brindabella Range bei Canberra (Eine pflanzensoziologischeStudie aus dem s~dostaustralischen Hart. Pub. Ak. der Wissenschaften. Franz Steiner Verlag GmBH. Wiesbaden. Mabbutt, J. A., 1968. In: Stewart, G. A. (ed.). Land Evaluation - papers of a CSIRO Symposium organized in co-op, with UNESCO, Canberra 26-31 August 1968. Owen, M. & Wyborn, D., 1979. Geology and geochemistry of the Tantangara and Brindabella 1:100 000 sheet areas, New South Wales and Australian Capital Territory. Dept. of National Development, Bureau of mineral Resources, Geology and Geophysics. Austral. Government Publ. Service, Canberra. Pook, E. W. & Moore, C. W. E., 1966. The influence of aspect on the composition and structure of dry sclerophyll forest on Black Mountain, Canberra A.C.T. Aust. J. Bot. 14: 223-242.
Pryor, L. D., 1938. The Botany, Forestry and Zoology of the Australian Capital Territory on an ecological basis (with map at 4.5 miles per inch scale). Commonwealth Government Printer Canberra. Pryor, L. D. & Moore, C. W. E., 1955. Plant communities. In: White, H.L. (ed.) Canberra, a nation's Capital, Part II, Halstead Press, Sydney. Rain J. R., 1920. Plan Shewing Millable Timber Paddy's River District Federal Territory Scale 20 chains per inch. Unpubl. Austral. Survey Office, Canberra. Thornthwaite, C.W., 1948. An approach toward a rational classification of climate. Geog. Res. 38 (1): 55-94. Ward, J. E. & Ingwersen, F., 1979. A checklist of vascular plant species in the Tidbinbilla Nature Reserve. Conservation Mem. 8. Dept. of the Capital Territory, Canberra. Whittaker, R. H., 1967. Gradient analysis of vegetation. Ecol. Rev. 49: 207-264. Williams, W. T., 1976. Other Ordination procedures. In: Williams, W.T. (ed.) Pattern analysis in agricultural science, CSIRO, Melbourne. Williams, W. T., Lambert, J. M. & Lance, G. N., 1966. Multivariate methods in plant ecology, V, Similarity analysis. J. Ecol. 54: 427-445. Wimbush, D. J. & Costin, A. B., 1973. Vegetation mapping in relation to ecological interpretation and management in the Kosciusko Alpine area. Div. Plant Ind. Tech. Paper 32. CSIRO Melbourne.
Accepted 8.9.1982.
179
Appendix Tidbinbilla community
Pryor's association
Costin's alliance
Story's communities
E. fastigata E. robertsonii
E. fastigata-E, viminalis E. fastigata-E, viminalis
E. fastigata-E, viminalis E. fastigata-E, viminalis
E. robertsonii-E, divesE. viminalis E. bridgesiana-E, dives E. dives E. rossii-E, macrorhyncha -E. mannifera E. dives-E, mannifera E. dalrympleana-E, dives
E. fastigata-E, viminalis
E. fastigata-E, viminalis
mountain community mountain community, gum community mountain community
E. dalrympleanaE. pauciflora E. pauciflora E. delegatensis
E. dives community E. fastigata-E, viminalis E. fastigata-E, viminalis E. dives community E. fastigata-E, viminalis E. macrorhyncha-E, rossii E. macrorhyncha-E, rossii E. rossii-E, macrorhyncha community E. macrorhyncha-E, rossii E. macrorhyncha-E, rossii E. dives community E. dives community E. delegatensis*E. delegatensisE. dalrympleana E. dalrympleana gum community E. delegatensis*E. delegatensisE. dalrympleana E. dalrympleana E. niphophila, E. pauciE. pauciflora-E, stellulata gum community flora-E, stellulata gum community E. delegatensis*E. delegatensisE. dalrympleana E. dalrympleana
'Best fit' relationship between nomenclature used in this work and that applied to adjacent areas. * E. gigantea in Pryor Pryor (1938) Pryor & Moore (1955) Costin (1954) Story (in CSIRO 1969)