Colonisation by ants of a planted bush garden plot in Perth, Western Australia

8 Pages·2002·1.7 MB·English

by  James Kain

COLONISATION BY ANTS OF A PLANTED BUSH GARDEN
PLOT IN PERTH, WESTERN AUSTRALIA
By JAMES KAIN, BRIAN E. HETERICK and JONATHAN D. MAJER
Department of Environmental Biology, Curtin University of Technology,
GPO Box U1987, Perth Western Australia 6845
ABSTRACT
A survey was undertaken of the ant fauna of a planted bush garden
(350 m2) at the City Beach Primary School, Perth, Western
Australia Although the bush garden had been planted with flora
native to the area, we found that its ant fauna most closely resembled
that of surrounding modified habitats (lawn, and lawn under trees)
rather than that of local woodland. Two cosmopolitan tramp ants,
present in large numbers in the bush garden, may have been
advantaged over many native ant species by local factors such as the
relatively small size of the bush garden, and the use of wood shavings
to line paths in the garden.
INTRODUCTION approximately 100m2), with additional
ground being added with each
Ants have been widely employed as subsequent stage The last planting was
indicators of environmental quality in 1994. The aim was to foster
because of their high diversity, the ease environmental awareness among young
with which they can be sampled and students enrolled at the school. The
identified, and their responsiveness to garden area measures approximately
changes in the environment (Majer 350m2, and has been planted or seeded
1983, Anderson 1990). The effects of with an estimated 34 species of plants,
mining (Majer and Nichols 1998), all of which are native to the Quindalup
pollution (Madden and Fox 1997), dune system associated with this region.
agricultural activity (Scougall ec al. However, at least one orchid species has
1993), invasion by exotic ant pests appeared in the garden, possibly from
(May and Heterick 2000) and old rootstock. Lawn surrounds the
urbanisation (Majer and Brown 1986) garden, and part of the lawn near the
have all been analyzed using ants as garden still has an overstorey of
biological indicators scattered Tuart (Eucalyptus
In 1989, an area of lawn at City Beach gomphocephala) and River Red Gum
Primary School, Perth, Western (Eucalyptus camaldulensis) trees.
Australia, was converted to a bush Between 25 April 2000 and 2 May
garden plot. Planting took place in four 2000, a survey was conducted examining
stages, with the first planting being the effectiveness of the bush garden as a
undertaken in May 1990 (in an area of haven for native invertebrates. At that
173
time the garden was found to have a diversity and ant species evenness. Ant
layer of scattered wood shavings, species richness was the total number of
originally provided for pathways species collected from each transect
through the garden The significance of Ant species diversity is here expressed
this is discussed below. For the purpose by the Shannon-Weaver diversity index
of analysis, ants were selected as a (FT), determined by the formula
surrogate for other invertebrate species FT = nlogn-X/jlcg/j/n
This report evaluates how closely the
where n is the total number of
ant fauna of the developing bush garden
specimens in the sample, and f (is the
approximates that of the original native
frequency of specimens of an individual
ecosystem the garden is designed to
species in the same sample. Ant species
recreate. We proceeded by comparing
evenness, also expressed as an index,
the ant fauna of the garden with that of
was determined by the formula:
(1) the nearby lawn, (2) the lawn with
Tuart overstorey, and (3) with that of J = H’l H’m
two areas of similar native bushland where FT is the logarithm of the total
that would formerly have existed at City number of species in the sample.
Beach Primary School. These two areas
Presence-absence data from each site
were situated in nearby Bold Park.
were then used for cluster analyses using
the SYN-TAX® computer package
(Podani 1995). Sorensens similarity
METHODS index was used in both analyses as it has
been shown to be one of the most
Five transects, each 20 m long, were
reliable measures for use with binary sets
measured out. One transect was
of data (Huhta 1979).
established in the lawn areas (L), one
in the lawn near the tuart trees (LT),
one in the bush garden (BG), and two
RESULTS
in separate areas (one in closed
woodland (WI) and one in a more A total of 2280 ants, representing 33
open aspect (W2)) in bushland in species, was collected in the pitfall
Bold Park. Ten pitfall traps were traps. Of these species, 9 were found
inserted in the ground an even to be undescribed or of uncertain
distance apart along each of these identity. In this report, therefore, each
transects. These traps were plastic of the latter has been assigned the
vials with a 41 mm internal diameter, JDM voucher number designating that
containing 25 ml of alcohol/glycerol particular species in the JDM Ant
(70/30 v/v). The traps were left open Collection. The bush transects (Wl
for seven days. and W2) produced the smallest
At the end of the trapping period the number of individual ants (27 (9
traps were emptied, and the ants sorted species) and 20 (20 species),
to species level using the Jonathan respectively), but had the highest
Majer ODM) Ant Collection in the species diversity (0.80 and 1.13) and
Department of Environmental Biology, species evenness (0.83 and 0.87)
Curtin University of Technology, as a indices. The bush garden transect
reference. Analysis of the ants was (BG) produced 696 (14 species), but
conducted employing the variables of the species evenness and species
ant species richness, ant species diversity indices were only 035 and
174
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
W1 W2
Transect
Figure L Dendrogram based on ant species presence/absence data illustrating the degree of
similarity of the four plots in terms of their ant community composition.
0.31, respectively. The lawn transect (454 ants) and 6.4% (35 ants), and for
(L) and lawn-with-trees transect (LT) LT, 32.8% (307 ants), 0.3% (3 ants)
showed a similar profile to BG (i.e, and 60.8% (569 ants). The percentage
551 individuals (13 species); 0.34 and occupied by these three species out of
0.30; and 936 individuals (13 species), the total ant fauna sampled was
and 0.42 and 0.38, respectively). The surprisingly similar across the three
above details are shown in Fig. 1. transects (i.e., 90.2% (BG), 89.2% (L)
and 93.9% (LT)). The sum of
When the species composition of the
catch was analyzed, the reason for the individuals of the remaining species
recorded from the transects comprised
high overall ant numbers from the
10% or less of the catch. In the two
transects set up in altered environments
bush transects, however, the individuals
(i.e., BG, L and LT) was obvious. Just
of no species comprised more than 20%
three ant species, lridomyrmex chasei
of the overall catch.
(native to the area), Tetramorium
bicarinatum (introduced) and The cluster analysis showed that the ant
Tetramorium simillimum (introduced), composition of the woodland transects
provided the bulk of the ant individuals W1 and W2 was similar, and quite
in these three transects. In BG, 1. chasei distinct compared with the other three
provided 65% (45 ants), T. bicarinatum transects. The similarity between
0.9% (6 ants), and T. simillimum 82.8% transects BG, L and LT was even more
(576 ants) of the overall catch. For L marked, with BG and LT being the most
the figures were 0.4% (2 ants), 824% similar (Fig. 2).
175
Table L Species and numbers of ants collected in the two bushland and three school plots
showing the overall ant abundance, species richness, species diversity and species evenness
index in the four areas
Ant taxon Bush transect 1 Bush transect 2
1 2 3 4 5 6 7 8 9 10 Total i 2 3 4 5 6 7 8 9 10 Total
Anonychomyrma itinerans perthensis (Forel) 1 1 3 3
Camponotus duilccus Crawley 0 0
Camponotus claripa sp. gp. (JDM 25) 0 1
Camponotus terebrans (Lowne) 0 13 13
Cardiocondyla nuda (Mayr)* 0 0
Crematogaster laeviceps chasei Forel 1 0
Dolichodtrrus clusor Forel 1 1 1 1
Iridomyrmexbickndli sp. gp. (JDM 14) 0 0
Jridomyrmex duuei Forel 6 1 7 3 1 4
Iridomyrmex rufoniger suchieri Forel 0 0
lridomyrmex sp. JDM 12 2 2 0
Mdophorns insularis Wheeler 0 1 1
Mdophorus sp. AN1C 3 0 1 1 2 4
Mdophorus sp. JDM 230 2 2 0
Mdophorus sp. JDM 783 0 1 3 4
M eranoplus rugosus Crawley 0 1 2 3
Monomoriumarenarium Heterick (in MSI 0 2 2
Monomorium leae Forel 0 1 1
M onomorium sydneyense Forel 0 1 11 2 14
Ochetellusglaber gp. (JDM 19) 0 1 2 1 4
Pachycondyla (Brachyponera) lutea (Mayr) 0 1 1
Paratrechina obscura (Mayr)* 0 0
Pheidolemegacephala (Fabricius)* 0 0
Pheidole sp. JDM 874 0 0
Rhytidoporurra levior Crawley 1 2 3 1 1 1 9 1 1 1 2 1 6
Rhytidoporurra inomata Crawley 1 2 3 0
Rhytidoponeraviolacea (Forel) 1 1 1 1 2
Solenopsis dark i Crawley 0 3 3
Stigmacrosbrouksi McAreavey 0 1 1
Stigmacros sp. JDM 115 0 1 1
Strumigenysperplexa (F. Smith) 0 0
Tetramoriumbicarinatum (Nylander)* 0 1 1
Tetramcrriumsimillimum (F. Smith)* 0 0
Ant abundance 1 10 7 3 1 0 1 0 2 2 27 5 18 8 3 2 7 3 0 20 4 70
Ant richness 9 20
Ant diversity 0.80 1.13
Ant evenness 0.83 0.87
176
Table 1 (cont.)
177
DISCUSSION with other native ant species,
particularly other Iridomyrmex.
Perhaps the most significant finding
from the survey was that the bush A second important finding was that no
garden ant fauna was highly similar to ant species dominated among the catch
the lawn ant fauna, and that from lawn from \V1 and W2 unlike the catch from
under trees. However, the ant fauna transects BG. L and LT. This is in
from all of these sites was substantially keeping with the findings of Majer
different to that of unmodified local (1977, 1978) and Majer et al. (1982)
environments. This has meant that six that ant evenness figures are typically
years of cultivating a relatively large high in lightly disturbed areas (such as
(350 m2) plot containing vegetation Bold Park, in this case) but can be very
native to the area has not resulted in an low in more highly disturbed areas. The
ant fauna similar to that of surrounding latter areas also frequently have very
native bushland. low plant diversity.
The two tramp ant species, Tetramorium Overall, the bush garden has, by and
bicarinatum and Tetramorium simillimum, large, failed to provide conditions
were a major component of the three suitable for the establishment and
modified sites, but were almost maintenance of native ant colonies.
completely absent (one individual of T. Admittedly, the rate of ant return to
bicarinatum in \V2) from the two bush rehabilitated sites is low in
transects. Tetramorium simillimum is a Mediterranean climates such as Perth
tiny species most frequently found in (Majer 1989, p. 165), but in this case
twig and leaf litter in gardens (B. the garden was still depauperate in
E. Heterick, pers. obs.). Consequently, native ant species at least six years after
it was no surprise that it was in the last planting. Nor is distance from
highest numbers in BG. Tetramorium bushland refugia containing native ant
bicarinatum, on the other hand, is a species an issue. City Beach Primary
well-defended medium-sized species School is less than five hundred metres
with a functional sting, and can often from Bold Park; well within reach of
be seen foraging in more open terrain alate, inseminated queens. The presence
such as lawn. The other major species, of scattered wood shavings throughout
the native odorous ant Iridomyrmex the garden may deter species that prefer
chasei, was present in both bush to found colonies in insolated, sandy
transects, but in low numbers (seven soil. The wood shavings may contain
and four individuals respectively). This chemical compounds that are toxic or
species occurred in low numbers in lawn repugnant to native ant species, but are
(L = 2 individuals), but was found in no barrier to the more adaptable tramp
larger numbers under trees (LT = 307) ants Invertebrate species suitable as ant
and in the bush garden (BG = 45). The prey may also be deterred from
presence of nectar, and of honeydew- inhabiting or visiting the garden bed
producing Hemiptera on shrubs and on because of the nature of the overlying
leaves and twigs of trees seems a wood shavings. Perhaps of greater
reasonable explanation for the higher significance is the fact that the
numbers of I. chasei in transects BG and dimensions of the garden may not be
LT. In W1 and W2 its lower numbers sufficient to exclude successful
could be explained by increased establishment of nests by queens from
competition for food and nest space the tramp ant colonies in the
178
surrounding lawn areas. Tramp ant ant communities to evaluate change. In:
colonies may then successfully kill Australian terrestrial ecosystems: a
queens of native ants landing in the review and a recipe Proceedings of the
garden, or outcompete nascent colonies Ecological Society of Australia 16: 347-
of native ants.
357.
The Coastal Brown Ant (Pheidole HUHTA, V. 1979. Evaluation of
megacephala), a major cause of local ant different similarity indices as measures
extinctions in the Perth regional area,
of succession in arthropod communities
has now become established in City
of the forest floor after clear-cutting.
Beach (B. E. Heterick, pers. obs.). An
Oecologia (Berlin) 41:11-23.
individual worker of this species was
MADDEN, K. E. and FOX, B. J. 1997.
collected from L. Should this ant
become established at the City Beach Arthropods as indicators of the effect of
Primary School, which seems likely in fluoride pollination on the succession
the near future, the ant fauna of the following sand mining. Journal of
bush garden and its surrounds will Applied Ecology 34:1239-1256.
become even more depauperate MAJER, J. D. 1977. Preliminary survey
Additional research to determine of the epigaeic invertebrate fauna with
whether native invertebrates, other particular reference to ants, in areas of
than ants, use the bush garden would be different land use at Dwellingup,
valuable Species that would otherwise Western Australia. Forest Ecology and
be absent from the area may be Management 1:327-334.
benefited by the presence of the garden,
MAJER, J. D. 1978. The important of
even if native ants are not.
invertebrates in successful land
In the meantime efforts should be made reclamation with particular reference to
by students and ground staff to ensure bauxite mine rehabilitation. In: (Ed. J.
that garden bed conditions, as well as
E. D. Fox) Rehabilitation of Miried Lands
the nature of the garden plants, should
in Western Australia workshop
approximate as closely as possible the
proceedings, Perth, Western Australia,
ground level environment of local
11 October 1978. Pages 47-61.
native bushland. Subtle, as well as
MAJER, J. D. 1983. Ants - useful
readily visible influences can affect the
bioindicators of minesite rehabilitation,
establishment of a healthy native ant
land use and land conservation status.
fauna. Attention to these influences
will, hopefully, enhance the Environmental Management 7:375-383.
establishment of native ants and other MAJER, J. D. 1989. Long-term
native invertebrates, as well as the birds colonization of fauna in reclaimed land.
and lizards that feed on them Should In: (Ed. J. D. Majer) Animals in Primary
Coastal Brown Ants manifest Succession/The Role of Fauna in
themselves, the use of Amdro®
Reclaimed Lands. Cambridge University
formulated bait will successfully regulate
Press, Cambridge. Part 2, Chapter 6 (pp.
these pests.
143-174).
MAJER. J. D. and BROWN. K. R. 1986.
The effects of urbanisation on the ant
REFERENCES
fauna of the Swan Coastal Plain near
ANDERSEN, A. N. 1990. The use of Perth, Western Australia Journal of the
179
Royal Society of Western Australia 69: fauna of the Perth metropolitan region,
13-17. Western Australia Pacific Conservation
MAJER, J. D. and Nichols, O.G. 1998. Biology &. 81-85.
Long-term recolonization patterns of PODANI. J. 1995. SYN-TAX 5.02.
ants in Western Australian rehabilitated Mac Computer Programs for Multivariate
bauxite mines, with reference to use as Data Analysis on the Macintosh System.
indicators of restoration success. Journal Scientia Publishing, Budapest, Hungary.
of Applied Ecology 35:161-181. SCOUGALL, S. A., MAJER, J.D. and
MAJER, J. D.. SARTORI, R, STONE, HOBBS, R.J. 1993. Edge effects in
R. and PERRIMAN, W. S. 1982. grazed and ungrazed Western Australian
Recolonisation by ants and other wheatbelt remnants in relation to
invertebrates in rehabilitated sand ecosystem reconstruction. In: (Ed. D. A.
mines near Eneabba, Western Australia Saunders, R. J. Hobbs and P. R.
Reclamation and Revegetation Research 1: Ehrlich) Nature Conservation 3:
63-81. Reconstruction of Fragmented Ecosystems.
MAY, J. E. and HETER1CK, B.E. 2000. Surrey Beatty &. Sons, Chipping
Effects of the coastal brown ant Pheidole Norton. Pp. 163-178.
megacephala (Fabricius), on the ant
180