NORTHERN

NATURALISJ

MUSEUM OF VICTORIA

29350

NT Field Naturalists' Club

1993 No.14

Northern Territory Naturalist 24 22 9 93)

NORTHERN TERRITORY
FIELD NATURALISTS' CLUB
- Founded 1977

Officers for 1992/93

President: Jenni Webber

Secretary: Greg Williams

Treasurer: Jeff Kum Jew

NT Naturalist Editors: Richard Noske
Alan Andersen

The objectives of the Northern Territory Field Naturalists' Club are to promote the
study and conservation of the flora and fauna of the Northern Territory. The club
holds monthly meetings and field excursions. Meetings are held in the Northern
Territory Museum of Arts and Sciences theatrette, Darwin, at 7:45 pm on the second
Wednesday of each month.

All members receive the bimonthly newsletter Nature Territory, and the annual
journal Northern Territory Naturalist. For information on membership rates, club
activities and publications please write to:

Northern Territory Field Naturalists' Club
PO Box 39565, Winnellie, NT 0821

Advice to contributors

The NT Naturalist publishes original papers treating any aspect of the natural
history of the Northern Territory or elsewhere in northern Australia. Contributors
to the NT Naturalist need not be members of the NT Field Naturalists' Club.
Submissions are considered on the understanding that thev are being offered for
publication solely by the NT Field Naturalists' Club, which will retain copy wright.
All manuscripts are refereed.

Manuscripts should be in duplicate, typewritten and double-spaced with generous
margins. To facilitate revision, the use of a word processor is strongly advised.
Authors should consult a recent issue of NT Naturalist for style and layout. Articles
should be written in a manner intelligible to readers without a specialist knowledge
of the subject. They should be appropriately illustrated with clear, black ink graphs,
diagrams or photographs. Reports of less than 500 words detailing observations of
something new or unusual in the field, for example on the behaviour of an animal,
will be published as Short Notes.

COVER: A group of waders (Bar-tailed Godwit, Great Knot, Mongolian Plover) at Lee
Point (A Hertog) ’

Northern Territory Naturalist 14 (1993)

1

Effect of large herbivore exclusion on understorey
biomass in three plant communities
on Cobourg Peninsula

D.M.J.S. BOWMAN

Conservation Commission of the Northern Territory, P.O. Box 496, Palmerston, NT 0831

Abstract

Gurig National Park on Cobourg Peninsula supports the only Australian feral herd of
Banteng Bos javanicus. Previous correlative surveys suggested that the feral Banteng
have a minor impact on the native vegetation in the Park. Three year experimental
exclusion of large herbivores at one site on the largely treeless coastal plains suggested
that grazing had no significant effect on the herb biomass compared to an adjacent
unfenced plot. However herb biomass in fenced eucalypt savanna and savanna-
rainforest ecotone communities was significantly greater than that in unfenced plots.
These results indicate that eucalvpt savanna and rainforest ecotone communities are an
important food source for large herbivores, and that the previously observed high
density of Banteng signs on the coastal plains may be unrelated to total food supply.

Introduction

Banteng Bos javanicus, a south-east Asian bovine, became feral on the Cobourg
Peninsula following the failure of a British settlement in the late 1840s. Unlike
buffalo Bubalis hubalus, the Banteng populations have grown slowly with an
estimated current population of about 3,500 beasts on Cobourg Peninsula, although
they occur at density greater than that reported from native habitats (Freeland
1990). For unknown reasons the Banteng have never spread far from their point
of introduction (Bowman 1993a).

A previous survey of twelve habitats on Cobourg Peninsula revealed that signs
and impacts of Banteng are focused on treeless coastal plains and monsoon
rainforests (Bowman & Panton 1991). Banteng are thought to graze preferentially
the treeless plains at night and retreat to the cover of the rainforest during the day
(Calabv & Keith 1974). Bowman & Panton (1991) found that Banteng signs were
least conspicuous in Eucalyptus savanna which comprises over95% of the vegetation
coverage on the Peninsula.

Bowman & Panton (1991) noted that their interpretations were based on correlative
evidence and suggested that firmer conclusions would require experimental

2

Northern Territory Naturalist 34 ( 3993 )

studies. Here I report the results of two field experiments. I describe the effect of
exclusion of large herbivore on: (1) herb biomass on treeless coastal plains; and (2)
herb biomass across the boundary between monsoon rainforest and eucalypt
savanna. Spotlight surveys and field traverses demonstrated that the primary
large herbivores at Cobourg Peninsula were Banteng, the remainder being
macropods.

PLATE 1 Dead Banteng in late dry season, Cobourg Peninsula (D Bowman)

Methods

Cover estimates were made at the start of the experiments because they are non¬
destructive, albeit less accurate than direct measurement of biomass. At the
conclusion of the experiment biomass was harvested from the plots for which
cover estimates had been made three years earlier. No attempt was made to
determine floristic changes because many species were unrecognisable at the time
of sampling.

Large herbivore exclusion on coastal plains

The vegetation on the coastal plains was dominated by the sedges Bulbostylis
barbarta and Fimbristylissp., the grasses Sorghum plumosum and Perotis rara and the
forbs Ptilotus amicus and Borreria sp. (see Bowman et al. 1990 for a description of

• 38

Northern Territory Naturalist 14 (1993)

3

this community). In an area of homogeneous vegetation on the coastal plains,
near the airstrip on Cobourg Peninsula, two 50 m x 50 m plots were established
spaced 50 m apart from each other. One plot was fenced with pigmesh wire to
exclude large herbivores. Fire breaks were graded around the perimeter of the
plots.

In June 1988 the percentage cover of vegetation was estimated for 25 0.5m x 1.0m
quadrats, using a modified cover abundance scale. The quadrats were
systematically set out 5m apart in a permanently marked 20m x 20m grid located
in the centre of each plot. In May 1991 the herb layer at each quadrat was
harvested to determine total herb biomass following oven drying of the samples
for 24 hours at 60°C.

Herbivore exclusion on a fire protected Eucalyptus savanna boundary
In 1988 two parallel 150m transects, spaced 100m apart, were established at right
angles to a monsoon rainforest boundary. The plant communities along these
transects comprised Eucalyptus savanna, rainforest ecotone and dry monsoon
rainforest as defined by Bowman (1993b). The herb layer in the Eucalyptus
savanna and rainforest ecotone was dominated by a variety of perennial grasses
including Sorghum plumosum, Heteropogon contortus, Chrysopogon latifalius, and
Sehima nervosa. Numbered steel pickets were placed every 5m along both transects.
A rectangular pigmesh fence was built around one transect. In June 1988 the total
percentage cover of herbaceous vegetation was assessed in two contiguous 1 m x
4m quadrats located at each 5m mark along the transects. In June 1991 the herb
layer from three 0.5m x 1.0m quadrats was harvested at each 5 m mark on each
transect to determine the herb biomass following oven drying at 600 C for 24
hours.

Data analysis

Non-parametric Wilcoxon tests were used to test the following null hypotheses:
(1) that there were no differences between cover of herbs on the two plots on the
coastal plains at the time of establishing the fencing experiment; (2) that three
years of large herbivore exclusion caused no difference in the herb biomass on
these plots; (3) that there was no difference in the cover of herbs in (savanna and
ecotone communities) on the rainforest boundary at the time of establishing the
fencing experiment. No comparison was made for the rainforest community as
there are no herbs on the forest floor; and (4) that three years of herbivore
exclusion caused no difference in the herb biomass in the savanna and ecotone.

Results

There was no significant difference (P > 0.05) between the fenced and unfenced
plots in the mean herb cover at the time of establishing the experiment (Table 1).

4

Northern Territory Naturalist 14 (1993)

After three years the mean herb biomass was not significantly different (P > 0.05)
between the unfenced and fenced plots (Table 1). There was no significant
difference (P> 0.05) between fenced and unfenced plots in the mean total cover of
herbs of either the eucalypt savanna or rainforest ecotone at the time of establishing
the experiment (Table 2). Mean total biomass was significantly greater on the
fenced transect than the unfenced transect after three vears of herbivore exclusion
(Table 2).

Discussion

This study shows that large herbivores have a significant impact on herb
biomass in eucalypt savanna and rainforest ecotone. One positive consequence
of this biomass removal may be a reduction of fire intensities on the rainforest
boundary.

TABLE 1 Mean (and standard error) of percentage herb cover in fenced and unfenced
plots on treeless coastal plains in 1988; and total herb biomass in 1991. Significance
of differences between plots for each year of measurement are determined by Wilcoxon
tests.

Fenced

Unfenced

Significance

Herb cover (%)

34 (1.9)

36 (1.5)

P > 0.05

Herb biomass (gm' 2 )

100 (9.3)

80' (7.7)

P > 0.05

n

25

25

The greater occurrence of Banteng signs on the coastal plains compared to the
eucalypt savanna (Bowman & Panton 1991) may be unrelated to total food supply
given that herbivore exclusion had no significant effect on herb biomass on the
plains. An alternative explanation for the conspicuous sign of Banteng on the
coastal plains, yet the apparent minimal impact of herbivory, is that grazing has
a stimulatory effect on herb growth on the plains, or stimulates the growth of
unpalatable species. Data on changes to species composition are required to
explore these ideas.

Northern Territory Naturalist 14 (1993)

TABLE 2 Mean (and standard error) of percentage herb cover in fenced and unfenced
eucah/pt savanna and ecotone on a monsoon rainforest boundary in 1988 (upper), and
herb biomass in 7992 (lower). Significance of differences between fenced and unfenced
communities are determined by Wilcoxon tests.

Fenced

Savanna

Unfenced Sig

Fenced

Ecotone

Unfenced Sig

Herb cover (%)

Mean

36

26 P>0.05

13

9

P> 0.05

SE

5.1

4.4

2.2

0.9

n

34

24

12

26

Biomass (gnr 2 )

Mean

132.9

76.8 P< 0.001

46.4

2.7

P< 0.01

SE

11.7

23.2

24.8

1.9

n

51

36

18

39

Assuming that the results demonstrate a large off-take (56 gnr 2 ) of Eucalyptus
savanna by Banteng then the biogeographic puzzle of why Banteng have not
exploited more of the Top End's savannas is further underlined (Bowman 1993a).
One possible reason for the geographic restriction of Banteng is that they must
supplement their diet of savanna herbs with minerals acquired from seawater.

The results presented here caution against extrapolation of herbivore impacts
from simple correlative studies of animal sign such as those of Bowman & Panton
(1991). More detailed studies are required to determine what the effect of Banteng
herbivory is on the ecology of Gurig National Park. These studies should ensure
that they overcome the deficiencies of this study such as the lack of replication of
exclosures and lack of data on the changes to different herb species biomass.

Acknowledgments

The Gurig Board is thanked for permission to conduct this research. Craig Martin,
Leonie McDonough, Bill Panton and Bruce Wilson are thanked for their field
assistance. Funding from the Endangered Species Unit, ANPWSand the National
Rainforest Conservation Programme supported some aspects of this research and
is gratefully acknowledged.

6

Northern Territory Naturalist 14 (1993)

References

Bowman, D.M.J.S. (1993a) Banteng. Aust. Nat. Hist. 24,16.

Bowman, D.M.J.S. (1993b) Establishment of two dry monsoon forest tree species on a fire-
protected monsoon forest-savanna boundary, Cobourg Peninsula, northern Australia.
Aust. /. Ecol. 18, 235-237.

Bowman, D.M.J.S., Panton, W.J. & McDonough, L. (1990) Dynamics of forest clumps on
chenier plains, Cobourg Peninsula, Northern Territory. Aust.). Bot. 38, 593-601.

Bowman, D.M.J.S. & Panton, W.J. (1991) Sign and habitat impact of banteng (Bos javanicus )
and pig (Sms scroja), Cobourg Peninsula, northern Australia. Aust. /. Ecol. 16, 15-17.
Calaby,J.H. & Keith, K. (1974) Mammals. In: Fauna survey of the Port Esslngton District, Cobourg
Peninsula, Northern Territory of Australia (eds. H.J. Frith & J.H. Calaby). CSIRO Div.
Wildlife Research, Melbourne, Tech. Paper No. 28, pp. 179-208.

Freeland, W.J. (1990) Large herbivorous mammals: exotic species in northern Australia. /.
Biogeog. 17, 445-49.

PLATE 2 Coastal calcareous plains typical of study site (D. Bowman)

Northern Territory Naturalist 14 (1993)

7

Assessment of Banteng Grazing on Coastal Plains,
Cobourg Peninsula: an Update on Bowman (1994)

W.J. PANTON

Conservation Commission of the Northern Territory, P.O. Box 496, Palmerston, NT 0831

Abstract

Late 1993 dry season herbage cover was found to be significantly different
between fenced and unfenced plots on the coastal plains at Smith Point,
Cobourg Peninsula. These findings contrast markedly with the results of
Bowman (1994) who found that at the same site there was no significant
difference in the biomass of herbs between the fenced and unfenced plots over
the three years (1988-1991) of his study. Comparison of Banteng Bos javanicus
scat counts on the plains for 1988 and 1993 indicate that there had been an
marked increase in the local density of Banteng at this site.

1 report these data to make the simple point that experimental studies must be
run for a sufficiently long time period to capture extreme or infrequent events,
which may be of greater ecological significance than results recovered for
"normal" years.

Introduction

Little is known about the impact of introduced ungulates (both feral and domestic)
on the Northern Territory savanna environment. Even basic data on the removal
of herbage is largely unquantified. Bowman (1994) found that, after three years
of exclusion of Banteng Bos javanicus from fenced plots on the coastal calareous
plains of Cobourg Peninsula, herbage biomass did not differ significantly from
that in adjacent unfenced plots, to which Banteng had access. However, over the
same time period, in the savanna-monsoon rainforest ecotone, significant
differences in herbage biomass were recorded.

Here 1 report differences in the cover of herbs between the fenced and
unfenced plots on the coastal plains at the end of the 1993 dry season. I
undertook this study because of the concerns expressed by Gurig National
Park rangers that utilisation of these plains by Banteng had increased
significantly. Unfortunately I could not examine the impact of Banteng grazing
at the monsoon rainforest boundary site because the exclosure used by
Bowman was destroyed by cyclone "Neville" in 1992.

8

Northern Territory Naturnlist 14 (1993)

Methods

In December 1993 a reassessment of herbage cover was undertaken for 25 pairs of
fenced and unfenced permanent subplots on the coastal sand plains at Smith
Point, Cobourg Peninsula. The percentage cover of herbs by species was visually
estimated by Braun-Blanquet cover classes. The plots had been harvested in 1991
by Bowman (1994), so cover reported here represents growth occurring over a 30
month period since harvesting. The cover scores were converted to percentage
mid-points and means and standard errors were calculated for the fenced and
unfenced plots. The effect of herbivore exclusion was determined by non-
parametric Wilcoxon Rank Sum tests.

Counts of scats on the sandplains were made to determine if the density of
Banteng scats was significantly different to 1988 densities (Bowman & Panton
1991). Scat density was determined from 25 quadrats, each measuring 10m x 20m,
placed 50m apart on two parallel transects. These transects were in approximately
the same location as those used by Bowman & Panton (1991). Wilcoxon tests were
used to compare the 1988 and 1993 Banteng scat data for these transects.

Results

The Wilcoxon test revealed a highly significant difference between the herbage
cover in the fenced and unfenced plots (P > 0.0001). The fenced plot had a mean
herbage cover of 9.2% (SE, 1.7%) and the unfenced plots had a mean cover of 0.7%'
(SE, 0.18%) cover. Only six non-grass species were encountered in sampled plots
and all except one, Cassythn filifonnis, were common to both fenced and unfenced
treatments.

Wilcoxon tests also revealed a highly significant (p > 0.0001) increase in the
density of scats from 162 (SE, 33) ha* 1 in 1988, to 526 (SE, 42) ha" 1 in 1993. It is
probable that the scats recorded on this survey were less than one year old, as
seasonal changes between wet and dry season and the extent of Banteng trampling
(Bowman & Panton 1991) on the sandplains would be conducive to the breakdown
rather than preservation of scats.

Discussion

The results of this study are at variance with those of Bowman (1994) who
concluded that there was no significant difference between the unfenced and
fenced plots, after three years of Banteng exclusion from the latter, on the coastal
plains at Smith Point. One likely explanation for these disparite findings is
increased grazing pressure on the plains from Banteng since Bowman's (1994)
study, as suggested by the greater density of scats in 1993. Aerial surveys in 1985
and 1989 (Bayliss & Yeomans 1989) detected some decrease in Banteng popula¬
tion, accounted for by off-take for disease control and trophy hunting; while

Northern Territory Naturalist 34 ( 3 993)

9

subsequent surveys in 1991,1992 and 1993 (Saalfield & Chatto 1994) indicating a
levelling of the population at around 1,200 to 1,800 head. These surveys detected
no change in the distribution of Banteng between broad scale sampling areas.

The scat results, however, suggest that Banteng density has been subject to a
substantial local fluctuation (assuming a direct relationship between herd size
and density of scats). The perception by the Gurig National Park Board of an
increased Banteng population in the Smith Point area, independent of this study,
was such that the Board allocated funds to fence the airstrip due to the increased
risk of light aircraft striking these animals when landing.

One possible cause of major fluctuations in Banteng numbers is movements due
to changes in food supply. Bowman (1994) reported significant changes in
herbage cover in the eucalypt savanna and savanna-monsoon rainforest ecotone
communities following Banteng exclusion, suggesting the importance of these
habitats in providing suitable forage for these mammals. Eucalypt savanna is the
dominant vegetation community of the peninsula (Wilson et nl. 1992), and is
capable of supporting annual fires. R. Chatto and K. Saalfield (pers. comm.)
observed extensive areas of intensely burnt savanna in the southern portions of
the peninsula during their aerial surveys of October 1993. These observations
contrasted with their impressions of a discontinuous fire pattern from previous
surveys. It is probable that many of the Banteng from these burnt areas of the park
were displaced to the unburnt north (e.g. Smith Point) in search of food.

This survey provides a cautionary note for the interpretation of studies on plant/
feral animal interactions. It is possible that occasional extreme events, like that
reported here, have greater deleterious effects on natural ecosystems than does
the low to moderate continuous pressure that occurs normally. Extreme, and
probably, irregular events of this type may help to explain rapid increases in soil
erosion or weed invasion within otherwise stable environments.

References

Bayliss, P. & Yeomans, K. M. (1989) Aerial survey of Buffalo, Cattle ami Bali Cattle in the Toy End
of the Northern Territory and Adjacent Areas. Report to the Northern Territory Department
of Primary Industries and Fisheries, BTEC Administration, November 1989.

Bowman, D.J.M.S. (1994) Effect of large herbivore exclusion on understorey biomass in
three plant communities on Coburg Peninsula. NT Naturalist 14, 1-5.

Bowman, D.J.M.S. & Panton, W.J. (1991) Sign and habitat impact of banteng (Bos javanicus)
and Pig (Sms scrofa), Cobourg Peninsula, Northern Australia. Aust. /. Ecol. 16, 15 -17.

Saalfield, K. & Chatto, R. (1994) Aerial survey of Bali cattle on Coburg Peninsula. Internal report
to the Gurig National Park Board.

Wilson, B.A., Brocklehurst, P.S., Clark, M.J. & Dickinson, K.J.M. (1991) Vegetation Survey of
the Northern Territory, Australia. Conservation Commission of the Northern Territory
Technical Report No. 49.

10

Northern Territory Naturalist 24 (1993)

Seasonal Patterns of Wader Populations in
Darwin, Northern Territory,
1974-1987

K.S. SHURCLIFF

P.O. Box 4878, Cairns, North Qld. 4870

Abstract

Counts of waders (Aves: Charadriformes) were made several times each month
from 1974 to 1987, in the Darwin vicinity. Northern Territory. Four species, all
breeding in Australia, were present year-round or were most numerous during the
dry season months. Most species (at least 28) were wet season migrants, arriving
and reaching maximum numbers in September or October. Approximately half of
these species showed a second influx in numbers during December and January.
Only a limited number of species showed a return passage through the study area.
Seasonal patterns shown bv the Darwin waders differ from those emerging from
other parts of northern Australia (e.g. Lane 1987), indicating that distinct regional
differences occur.

Introduction

Darwin, on the north coast of the Northern Territory, is one of the 20 most
important sites for aggregations of waders (Aves: Charadriformes) within Australia
(Lane 1987). It is the closest of these sites to south-east Asia and the Indonesian
archipelago, and thus studies of waders near Darwin could play a prominent role
in increasing our understanding of wader migration throughout this part of the
world. Published records to date for this area are based upon limited sampling,
such as reported in Lane (1987), or mostly unquantified observations, such as
reported by Crawford (1972) for the period 1967-71.

Darwin has a marked wet-dry monsoonal climate, with rainfall between years
being highly variable (Ridpath 1985). The wet season usually extends from
December-March, the dry from June-September, with short intervening transition
periods. The studies reported here commenced in 1974 and continued until mid-
1987. They form part of more extensive records kept for all bird species in the area.
Previous reports for the study area have been made for other groups of birds
(McKean 1981,1986; Thompson 1982,1984).

Methods

Surveys were made several times each month, at each of six habitat types: beach
and intertidal mudflat (Lee Point and Buffalo Creek), mangroves (Buffalo Creek,

Northern Territory Naturalist 24 (1993)

11

Palmerstonand Sanderson sewage ponds), sewage ponds (Sanderson, Palmerston),
saltflats (Leanyer), estuarine swamp merging into freshwater swamp (Holmes
jungle Swamp), and open grassland (1 lolmes Jungle, Lee Point road). A total of
1,330 surveys were made. The maximum count fora species for any given month
was used in the data analysis. This was used rather than all the individual counts
per month, as variations across surveys were high due to differences in time and
stage of tide, observers and length of time per survey. The time period spent on
each survey was usually about 60 minutes, but varied between 30 and 180
minutes.

Because the data were not standardised, it is assumed that the numbers represent
relative data only; therefore, non-parametric tests of statistical significance were
applied. To test whether the population samples were significantly different from
month to month, the data were analysed using the non-parametric Friedmann
two-way analysis of variance. The monthly figures were ranked within a given
year from lowest to highest. The test determines whether the rank totals for the
months over the 14 years differ significantly.

PLATE 3 Large Sand Plovers and SanderUngs, Darwin (A. Hertog)

Results

Thirty-three wader species were recorded regularly during the study. There were
also several records of vagrant species, but as these have been documented
elsewhere (e.g. McKean el al. 1976; McKean & Hertog 1981), and do not show
population trends, they are not dealt with in this paper. Twenty-six species
occurred frequently enough to test for significant changes in population levels.

12

Northern Territory Naturalist 34 (1993)

These fell into three groups based on patterns of abundance through the year:
those which were present year-round without marked seasonal changes (one
species), those which showed dry season peaks (three species), and the remainder
which showed wet season peaks (22 species).

The only species that showed no significant seasonal changes in population
was the Red-capped Plover Charadrius ruficapillus, although January showed
consistently larger numbers (Fig. 1). The three species which were most
numerous during the dry season months of June-September were the
Australian-breeding Masked Lapwing Vancllus miles, Black-winged Stilt
Himantopus himantopus, and Australian Pratincole Stiltia isabella (Fig. 1). The
remaining species were long-distance migrants from the Palaearctic region
which showed highest numbers during the wet season, or wet season transition
months (October-November) (Fig. 2).

RED-CAPPED PLOVER

BLACK-WINGED STILT

MASKED LAPWING

AUSTRALIAN PRATINCOLE

L n
C

3

O

W

CD

(/)

c

3

*3

CD

7T

FIGURE 1. Mean counts of individuals per month for four Australian-breeding species.
Except for Red-capped Plover, graphs show both mean counts of individuals (-») and sum
of monthly ranks () over the full study period, 1974-87.

Northern Territory Naturalist 14 (1993)

13

GREY PLOVER

LESSER GOLDEN PLOVER

LARGE SAND PLOVER

RUDDY TURNSTONE

WHIMBREL

EASTERN CURLEW

LITTLE CURLEW

FIGURE 2 Mean counts of individuals ( -a-) and sum of monthly ranks ( ) over the

full study period , 1974-8 for wet season migrant waders.

Sum of Ranks Sum of Ranks Sum of Ranks Sum of Ranks

14

Northern Territory Naturalist 74 (7993)

WOOD SANDPIPER

GREY-TAILED TATTLER

COMMON SANDPIPER

MARSH SANDPIPER

BLACK-TAILED GODWIT

GREENSHANK

TEREK SANDPIPER

FIGURE 2 (continued)

Northern Territory Naturalist 14 (1993)

15

Arrival of Long-distance Migrants

Long-distance migrant species generally arrived in Darwin during September, but
maximum numbers were in either September or October, with the exception of
Oriental Pratincole Glareola maldivarum (Fig. 2). Based on the measure of the total
ranks, those species with initial peak numbers in September (Group 1) were Grey
Plover Pluvialis squatarola, Mongolian Plover Charadrius mongolus , Ruddy Turnstone
Arenaria interpres, Whimbrel Numenius phaeopus, Grey-tailed Tattler Tringa brevipes,
Terek Sandpiper T. terck, Bar-tailed Godwit Limosa lapponica, Great Knot Calidris
tenuirostris, Curlew Sandpiper C. ferruginea and Sanderling C. alba.

Other long-distance migrant species, although they commenced arriving in
September or earlier, peaked later, in October (Group 2). These included Large
Sand Plover Charadriusleschenaultii, Little Curlew Numenins minutus, Greenshank
Tringa nebularia, Marsh Sandpiper T. stagnatilis, Black-tailed Godwit Limosa
limosa, Sharp-tailed Sandpiper Calidris acuminata and Red-necked Stint C. ruficoltis.
Other species (Group 3) were not consistent in their arrival month, including
LesserGolden Plover Pluvialis dominion, Eastern Curlew Numenius madagascarknsis,
Wood Sandpiper Tringa glareola and Common Sandpiper T. hypoleucos.

December-]anuary influx

About a half of the long-distance migrants showed a second influx during the
months of December and January. After building up numbers during September
and October, these species showed substantial decreases during November and
early December, but subsequently increased again in late December and January.
Their numbers characteristically declined again in February. This December-
January influx was found in all the species of Group 1 (September maxima), as
well as for the Large Sand Plover, Common Sandpiper, Black-tailed Godwit and
Red-necked Stint. Two other species, the Greenshank and Marsh Sandpiper,
showed a similar pattern, but with the increase in February.

Departure and return passage of long-distance migrants

Most long-distance migrants (16 species) departed during March and early April,
though three species departed in late April-May. Only one species, the Large
Sand Plover, showed strong evidence of a return passage through the study area
(presumably from southern Australia), with substantial increases in the study
area during February-April. Other species showing minor increases during this
period include Mongolian Plover, Sanderling, Whimbrel, Eastern Curlew, Ruddy
Turnstone and Greenshank.

Several long-distance migrants were present in limited numbers throughout the
northern hemisphere breeding season, June-July. These were Great Knot,
Greenshank, Whimbrel, Large Sand Plover, Mongolian Plover, Red-necked Stint
and Eastern Curlew.

16

Northern Territory Naturalist 14 (1993)

GREAT KNOT

SHARP-TAILED SANDPIPER

in

c

3

o

*3

3

7T

V)

SANDERLING

FIGURE 2 (continued)

ORIENTAL PRATINCOLE

LT>

C

3

o

73

BJ

3

pr

Restricted Season Species

LittleCurlew Nwnenius tninutns and Oriental Pratincole showed discrete, restricted
seasonal patterns, occurring only from September to January (Fig. 2). Oriental
Plover Clumuirius veredus, which occurred in very small numbers (see below),
showed a similar trend. All three species feed on open grasslands, but move out
of the study area after the start of the wet season, when local feeding grounds
become flooded and alternative areas become suitable due to widespread rains.
Little Curlew disperse inland after the onset of rains flood their feeding grounds
and do not show a return passage through the study area (McKean ct nl. 1986),
probably due to lack of suitable feeding grounds at that time.

Northern Territory Naturalist 14 (1993)

17

Uncommon Species

The remaining species occurred in numbers too low to show statistically significant
population patterns. A few Beach Thick-knees Burhinus neglectus were regularly
recorded, except from April to July. Pied Oystercatcher Haematopus longirostris
and Sooty Oystercatcher H. futiginosus were found in every month. Several other
species that breed within Australia were absent during at least part of the wet
season. Black-fronted Dotterel Charadrius rttelanops and Red-kneed Dotterel
Erythrogonys ductus were absent from November and December respectively, to
February. Red-necked Avocet Recurvirostra novaehollandiae was never recorded
from November to April. These absences could coincide with breeding elsewhere.
Their abundance in the study area varied substantially from year to year.

Other uncommon species showed patterns similar to the more common wet
season migrants. Swinhoe's Snipe Gallinago megain was present from October to
April, while Little Ringed Plover Charadrius dtthius occurred only from September
to March. A few species apparently disappeared, and then reappeared later
during the wet season. Oriental Plover was found from August to January, and
subsequently in April, on return passage. Red Knot Caladris canutus was present
in September-October, December-January, and March-May.

Discussion

Australian breeding species were the only ones which showed higher numbers
during the dry season, possibly representing influxes of birds from further sou th,
where conditions would be drier and colder. The large numbers of Red-capped
Plover recorded during January in Darwin conflict with the data of Crawford, as
reported in Lane (1987), which indicated that highest numbers occurred during
the dry season on saline plains. This pattern clearly does not hold for the range
of habitats surveyed in this study.

Most migrants arrived during the period August-October; 17 species being very
consistent in their arrival month. Approximately 50% of maximum counts
occurred during September, 30% in October and 20% in August, with one species
(Oriental Pratincole) having maximum numbers in December. This timing
appears to be slightly later than the maximum counts found for north-western
Australia (see Lane 1987). It suggests that the migrants landing in the north-west
may originate from separate places in Asia to those landing in Darwin; or that
Darwin birds have moved east after first arriving in north-west Australia, as has
been suggested by Lane.

This study found a second influx in numbers for several long-distance migrants
during Decemberand January. Two waves of migration had previously onlybeen
reported for Red-necked Stint (Lane 1987). The second influx could represent
birds newly arriving from overseas (perhaps juveniles, M. Barter pers.comm.) or

ERRATA: Northern Territory Naturalist 14 (1993)

1. On p 7 and back cover:

Assessment of Banteng Grazing on Coastal Plains, Cobourg Peninsula: an Update on Bowman (1994)

should read:

Assessment of Banteng Grazing on Coastal Plains, Cobourg Peninsula: an Update on Bowman (1993)

2. On pp 7, 8, 9

Bowman (1994) should read: Bowman (1993)

3. On p 9 in References

Bowman, D.J.M.S. (1994) Effect of large herbivore exclusion on understorey biomass in three plant communities.

should read:

Bowman, D.J.M.S. (1993) Effect of large herbivore exclusion on understorey biomass in three plant communities.

4. On back cover (contents)

Seasonal patterns of wader populations in Darwin, Northern Territory, 1974-1987. K.S. SHURCLIFF and H.A.F. THOMPSON
should read: K.S. SHURCLIFF

18

Northern Territory Nntnrnlist 14 (1993)

internal migration, especially from the north-western Australia. Unfortunately
no data are available on the age of birds in this study. Many of the species involved
are those which spend most of their time in Australia in the north (e.g. Large Sand
Plover, Great Knot, Black-tailed Godwit), but also include some predominantly
southern species according to Lane (1987), e.g. Red-necked Stint and Curlew
Sandpiper.

Only seven long-distance migrant species showed evidence of a return passage
through the area. As individuals or groups of only seven Palaearctic-breeding
species remained through their breeding season, the Darwin area does not appear
to be an important overwintering site, such as the Gulf of Carpentaria.

Acknowledgements

Most (90%) observations reported in this paper were made by the author, H.
Thompson and J. McKean. Other observers included K. Fisher, L. Fisher, J.
Estbergs, F. van Gessel and A. Hertog. Very constructive suggestions on an
earlier draft of this paper were made by M. Barter, R. Jaensch and R. Noske.

References

Crawford, D.N. (1972) Birds of the Darwin area, with some records from other parts of the
Northern Territory. Emu 72, 131-148.

Lane, B. (1987) Shorebirds in Australia. Nelson, Melbourne.

McKean, J.L. (1981) The status of gulls and terns (Laridae) in the Darwin area, N.T. 1974-
1980. Aust. Seabird Croup Neivsl. 15, 11-16.

McKean, J.L. (1986) The status of cuckoo-shrikes (Campephagidae) in the Darwin area,
Northern Territory, 1974-1984. NT Naturalist 9, 8-14.

McKean, J.L. & Hertog, A.L. (1981) Some further records of uncommon migrant waders
near Darwin, N.T. NT Naturalist 4, 10-12.

McKean, J.L., Thompson, H.A.F. & Estbergs, J.A. (1976) Records of uncommon migrant
waders near Darwin, N.T. Aust. Bird Watcher 6, 143-148.

McKean, J.L.,Shurcliff, K.S. & Thompson, H.A.F. (1986) Noteson thestatusof LittleCurlew
Numenius minutus in the Darwin area, 1974-1985. Aust. Bird Watcher 11, 258-260.

Ridpath, M.G. (1985) Ecology in the wet-dry tropics: how different? In: Ecology of the Wet-
Dry Tropics (ed. M.G. Ridpath). Proc. Ecol. Soc. Aust. 13, 3-20.

Thompson, H.A.F. (1982) The status of cuckoos (Cuculidae) in the Darwin area. Northern
Territory, 1974-1980. NT Naturalist 5, 13-19.

Thompson, H.A.F. (1984) The status of kingfishers and their allies (Coraciiformes) in the
Darwin area, N.T. NT Naturalist 7,18-29.

Northern Territory Naturalist 14 (1993)

19

Temporal Dynamics of two Tropical Ant Species,
Iridomyrmex sanguineus and Oecophylla smaragdina,
in relation to Temperature and Humidity

A. M. LANE

CSIRO Division of Entomology, PMB 44, Winnellie, NT 0821.

Abstract

The activity of the tropical meat ant Iridomyrmex sanguineus, in the Darwin region,
is strongly affected by environmental temperature and relative humidity. Activity
peaked at air temperatures 3mm above the soil surface between 30° and 35°C and
was limited at temperaturesexceeding38°C with a relative humidity less than 62%.
By contrast, the green weaver ant Oecophylla smaragdina, did not adjust its activity
in response to changing air temperatures experienced in the months March to May.
It did, however, appear to respond to light, being more active at night than during
daylight over the study period. It is unlikely that weaver ants in their preferred
habitat would ever experience temperatures high enough to cease activity.

Introduction

Ants often follow distinct daily cycles of activity, and operate within limits of
temperature and humidity (Holldobler & Wilson 1990). They tend to be strongly
thermophilic, with few species foraging at all below 10°C (Holldobler & Wilson
1990). Some species can tolerate extremely high temperatures: for example, the
central Australian ant Melophorus bagoti remains active during the hottest period
of a summer day when ambient temperature exceeds 50 n C (Christian & Morton
1992). Many species adjust their foraging activity in response to several factors
including ambient and surface temperatures. In tropical habitats, daily and
seasonal changes in foraging activities of ants are related to both temperature
(Torres 1984) and moisture (Levings 1983). This study uses field observations to
assess the effects of ambient and surface temperatures, and relative humidity, on
foraging activity in two species of tropical ants, Iridomyrmex sanguineus and
Oecophyl la sma ragd ina.

Materials and methods

The northern meat ant, Iridomyrmex sanguineus (Formicidae: Dolichoderinae), is
a relatively large (length 6-8 mm), dominant, diurnal ant species (Greenslade

20

Northern Territory Naturalist 14 (7993)

1976; Andersen 1992), very common in open vegetation in tropical Australia.
Meat ants generally live in large colonies in underground nests (Barker &
Greenslade 1982), with each nest having one to many entrance holes, each of
which leads to a separate system of galleries and chambers (Greenslade 1979).
A nest may be occupied by the same population for many years (Greenslade
1975).

The green weaver ant, OecophyHa smaragdina (Formicidae: Formicinae), is common
throughout tropical Australia and South East Asia. It typically lives in large
colonies and constructs nests by weaving the leaves of trees, which are bound by
silk produced by the larvae. Green ants are one of the most successful ant species
due to their ability to form large colonies and their aggressive and territorial
behaviour. A colony may dominate several trees at one time and populations of
half a million often occur (I lolldobler & Wilson 1990).

The present study took place near Darwin (12°26' S, 130°55' E) during March-May
1993, the transition period between wet and dry season, when conditions are often
overcast with irregular heavy rain storms. The air temperature ranges from about
22° to 35°C, and relative humidity from about 35 to 98%. The meat ant nest was
located on the grounds of the CSIRO laboratories Berrimah, 10 km north east of
Darwin. The nest was in bare red earth and had 11 entrances, roughly in a line
with a maximum separation of 23 m. Only one hole, centrally located, was
selected for monitoring. The green ant nest was located in Nightcliff, 10 km north
of Darwin. The monitored nest was one of eight located in a lemon tree, and was
approximately 2 m above ground level on the perimeter of the tree. It had three
openings but only the main entrance hole was monitored.

Activity in both species was measured by counting the number of ants entering
and leaving one entrance over five minute periods. Entry/ exit counts were
combined to give a measure of activity. For /. sanguineus the following variables
were measured at each count: air temperature at 2 m (Ta), ground surface
temperature (Tg), temperature at 3 mm above ground ("ant height") (Tah), nest
temperature at 20 cm below surface (Tn), relative humidity at 2 m (Rh). Tg and
Tah readings were taken approximately 1 m from the monitored hole. For O.
smaragdina, only air temperature (at 2 m), nest temperature and relative humidity
at 2 m were measured. Temperatures were measured with thermocouples, the
nest probe being left in the ground for the entire study, and relative humidity was
measured with a sling psychrometer.

To determine which environmental variable most affects the activity of /.
sanguineus, all subsets regressions of the variables against total activity were
carried out, using the adjusted r 2 as the selection factor. A paired t-test was used
to test the null hypothesis that the nest temperature of O. smaragdina was similar
to ambient temperature.

Northern Territory Naturalist 74 ( 7 993)

21

Results

Iridoim/rmex sanguineus

The range of air and nest temperatures were much narrower than soil surface or
ant height temperatures (Fig. 1). The model of activity with the variables Rh, Tah
and Tah 2 had the highest adjusted r 2 (Table 1) and demonstrates that /. sanguineus
prefers high humidity and medium temperatures (Fig. 2). Activity was also
significantly related to ground temperature (r 2 = 0.5181, p = 0.0000) and air
temperature (r 2 = 0.3297, p = 0.0000). Activity was lowest at low air temperatures,
and at high air temperatures with low humidity and generally, minimum activity
occured at the extreme temperatures in the ranges.

TABLE 1 Summary of the best model describing activity of Iridomyrmex sanguineus
over five minute periods

Variable

Adjusted r 2

P

Rh

0.2323

0.0004

Rh - Rh 2

0.3939

0.0000

Tah

0.2479

0.0003

Tah - Tah 2

0.4304

0.0000

Rh + Tah - Tah 2

0.5272

0.0000

At the cooler temperatures, the ants adopted a crouched position, moved very
slowly and stayed in close proximity to the nest. Some remained stationary for
short periods of time. At higher temperatures their bodies were raised from the
ground and they moved very quickly around and away from the nest.

O. smaragdina

The activity of O. smaragdina was unrelated to air temperature (r 2 = 0.0392; Fig.
3a) and relative humidity (adjusted r 2 = 0.0270). Time of day however, was a
significant factor with an r 2 of 0.5488 (Fig. 3b). Maximum activity occurred
during the dark hours (before 06:00 and after 18:00) and showed a 507c
decrease during the daylight hours (after 06:00). There was no significant
difference between mean air temperature and mean nest temperature (f = 0.47,
d.f. = 59, P > 0.05).

22

Northern Territory Naturalist 14 (1993)

60 -I

50

40

o

o

0)

aj 30

CD

Q.

E

a>

I—

20

10 -

Air Soil surface Ant height Nest

(Ta) (Tg) (Tah) (Tn)

FIGURE 1 Range of the temperature, means and standard errors at different locations.
The boxes indicate the range between the 25th and 75th percentiles of the data; the lines
mark the 50th percentiles, and capped bars indicate the 10th and 90tli percentile points.

T

Discussion

The results suggest that 1. sanguineus activity is strongly a ffected by environmental
temperature and relative humidity. The optimum foraging ant height temperature
was around 35°C. The high temperatures occurred mainly after 11:30 and low
temperatures occurred before 07:00 hours.

By contrast, O. smaragdina was not strongly affected by air temperature, at least
within the range of 24° to 37°C, or relative humidity. Because these ants spend the
majority, if not all, of their time in trees, they would be well shaded from direct
and reflected solar radiation and would not experience the high temperatures
occurring at ground level. Also, the thermal conductance of vegetation is lower
than that of bare ground so the risk of desiccation induced by activity at high
temperatures would be considerably reduced. The highest air temperature
recorded was 37.5°C, which is above average, but the activity of green ants at this
temperature was not significantly different than at some lower temperatures. The
highest air temperature ever recorded in Darwin was 38.9°C in October 1991

Northern Territory Naturalist 74 (1993)

23

(Bureau of Meteorology, Darwin, pers. comm.) so it seems unlikely that green ants
would ever experience air temperatures sufficiently high in the Darwin region to
prevent their activity. Greenslade (1972) found that this species is well adapted
to tropical environments and is tolerant of variations in both temperature and
humidity.

Ground temperature and temperature at ant height for I. sanguineus varied
considerably more than air and nest temperature (Fig. 1). Both fluctuated readily
with slight changes in wind speed and cloud cover. The ants were active over a
large range of environmental temperatures but for a given temperature, activity
was generally greater at high humidity when the risk of desiccation would be
considerably reduced. It is significant that the lowest ant count of only one over
a five minute period occurred at the highest ground temperature of 54.2°C when
relative humidity was only 45%. Nest temperature by contrast, showed very little
variation with a mean of 33.6°C, ranging from 26° to 36°C. It rarely fell below Ta,
and then only slightly. Ground-dwelling insects intercept direct and reflected
solar radiation as well as infrared radiation, conductive and convective heat from
the ground. Therefore, they are particularly prone to desiccation (Withers 1990).
When ambient temperatures are high it is imperative that they have some means
of avoiding overheating and, for meat ants, a relatively cool nest environment
provides a thermal refuge.

FIGURE 2 Actual and fitted values of the activity of I. sanguineus over five minute
periods in relation to temperature 3 mm above the soil surface (Tali) and relative humidity
(Rh). The equation for the model is, Activity = 14.6Rh + 314Tah - 4.2Tah 2 - 6350.

24

Northern Territory Naturalist 14 (1993)

'>

u

<

200 - (a)

150 -

100 -

50 1

0 J ■ . . . . .■——I

20 25 30 35 40

Temperature (°C)

200 -| (b)

’>

u

<

150 -

100 -

50 -

O

Oo 0

• O

Oo

oo

0 J i i i i i i i i i i i i i i i i i i i

<;oo sou inno 1200 1400 i«oo isoo 2000 2200

Time of day (hour)

FIGURE 3 Menu activity of] O. smaragdina over five minute periods at (a) air
temperature and (h) time of day (open circles = day; closed circles = night).

Northern Territory Naturalist 14 (1993)

25

It is possible that the energy available, if any, at low ambient and surface
temperatures is too low for effective foraging. The low body posture adopted by
I. sanguineus would maximise heat gain from the boundary layer of the ground
surface. The elevated posture adopted at high ground temperatures would raise
the ant's body into the cooler part of the thermal boundary layer and increase
convective heat loss (Withers 1990). Melophorus bagoti spends up to 50% of the
foraging time during midday in the refuge of sparse vegetation, and, despite soil
surface temperatures exceeding 70°C, there is never a cessation of activity
(Christian & Morton 1992). It is possible that 1. sanguineus also uses this strategy
for thermal respite. The fact that no other ant species were observed around the
nest site and that meat ants are relatively large and aggressive, suggests that their
foraging times are not affected by interspecific competition.

The nest temperature of O. smaragdina fluctuated with air temperature so it would
not provide a thermal refuge as it does for I. sanguineus. When the sun was shining
directly on the nest of O. smaragdina, the majority of ants spent most of their time
in the shade on the underside of the nest, while at other times, particularly in the
early morning, they moved over the entire nest (S. Walsh pers.obs.). No
thermoregulatory behaviour in response to thermal extremes was observed. The
lowest temperature recorded was 24.8°C and it was at this temperature that the
highest activity occurred. This reading was taken at night. Activity appears to be
limited by light, peaking at and after sunset and before sunrise. Activity
decreased by approximately 50%' during the daylight hours. It may have been
useful to take more readings during the night to determine if this trend is
consistent as not all nests or colonies may exhibit nocturnal behaviour.

This study revealed that I. sanguineus and O. smaragdina respond quite differently
to changes in environmental temperature. The activity of the meat ant was
limited by environmental temperature extremes and low relative humidity, while
the green ant was tolerant of changes in ambient temperature, responding more
to light. Further observations on other nests of 1. sanguineus and O. smaragdina ,
and at other times of the year, are required to determine the generality of these
findings.

Acknowledgements

I thank L. Knight and S. Walsh for carrying out the field work on O. smaragdina.
1 also thank W. M. Lonsdale and A. N. Andersen for guidance and helpful
comments on a draft of the manuscript.

References

Andersen, A.N. (1992) Regulation of "momentary" diversity by dominant species inexceptionallv
rich ant communities of the Australian seasonal tropics. Am. Nat. 140, 401-420.

26

Northern Territory Naturalist 24 (1993)

Barker, W.R. & Greenslade, PJ.M. (1982) Evolution of the Flora and Fauna of arid Australia.
Peacock Publications, South Australia.

Christian, K.A. & Morton, S.R. (1992) Extreme Thermophilia in a Central Australian Ant,
Melophorus bagoti. Physiol. Zool. 65, 885-905.

Greenslade, P.J.M. (1971) Phenology of three ant species in the Solomon Islands. /. Aust.
Ecology 10, 241-252.

Greenslade, P.J.M. (1972) Comparativeecologyoffourtropicalantspecies. InsectesSociaux
(Paris) 19,195-212.

Greenslade, P.J.M. (1975) Short term change in a population of the meat ant Iridomyrmex
purpureas (Hymenoptera: Formicidae). Aust. /. Ecology 23, 511-522.

Greenslade, P.J.M. (1979) A Guide to the Ants of South Australia. South Australian Museum.
Holldobler, B. & Wilson, E.O. (1977) Weaver Ants. Scientific American 237,146-154.
Holldobler, B. & Wilson, E.O. (1990) The Ants. Harvard University Press, Massachusetts.
Withers, P.C. (1991) Comparative Animal Physiology. Sanders College Publishing, Orlando,
Florida.

PLATE 4 The northern meat ant, Iridomyrmex sanguineus (B. McKaige)

Northern Territory Naturalist 14 (1993)

IT

SHORT NOTES

The Use of Small Pebbles in the Web Retreat of a
Comb-footed Spider, Achaearanea sp.
(Araneae: Theridiidae)

The comb-footed spiders of the family Theridiidae typically construct a three-
dimensional tangle web, or cob-web, often in a sheltered site. A retreat where the
spider resides is sometimes made towards the centre of the web. This retreat is
usually conical in shape (with the point uppermost) if constructed wholly of silk,
but the shape may be altered if other building materials are incorporated.

In December 1991 a number of Achaearanen sp. spiders were observed establishing
themselves under the window-sill of a recently constructed house at Humpty
Doo, Northern Territory. The window had a northerly aspect and was protected
by the roof of a verandah, approximately 3.2 m wide. The webs of the spiders were
similar to other theridiid webs observed by the author, albeit rather sparse. The
construction of the retreat, however, was unusual. A number of tiny pebbles were
incorporated into the silk of the retreat, in some cases packed quite tightly. Similar
pebbles were found scattered over a section of the paved floor of the verandah,
one metre below the web site. The pebbles were up to 3 mm in diameter, often
larger than the spiders' abdomen.

As the months passed, and as other webs and retreats appeared, the incidence of
pebbles as the major component of retreats decreased. In order to establish
whether the use of the pebbles was dependant on the type of material available,
two spiders were removed from their "pebblecrete" retreats, and placed in
separate jars. The floor of the jars were covered in material gathered from below
the web sites, mostly small pebbles, sand grains and pieces of dry leaves,
presumably blown in from the yard. Within several days the reconstruction of the
retreats had begun. The spiders used the smaller building materials such as the
sand and leaf components rather than the larger, heavier pebbles.

Subsequently 1 found a new web and retreat, which was built directly over an
ants' nest, and composed of sand grains deposited by the ants. Other retreats
observed nearby in pot-plants were constructed from fragments of potting
mix. These observations suggest that the original pebble retreats were
constructed using the only material available at the time. With the availability
of a wider variety of building material in the immediate area, the occurrence
of the larger pebbles in the retreat decreases. Use of such heavy building

28

Northern Territory Naturalist 14 (1993)

materials must have a considerable energetic cost, indicating a high premium
for unreinforced retreats.

I thank Dr Mike Gray (Australian Museum) for identification of the specimen, and
Drs Mark Harvey (WA Museum), Alice Wells (ABRS) and Richard Noske (NTU)
for comments on the text.

JENNI WEBBER, Museum & Art Gallery of the Northern Territory, Darwin, GPO Box
4646, Darwin, N.T., 0801

* this article was originally published in Australasian Arachnology, no. 45.

Observations of the Port Darwin Sea Snake
Hydrelaps darwiniensis

The Port Darwin Sea Snake Hydrelaps darwiniensis (Hydrophiidae), was
named from two specimens collected near Darwin (Boulenger 1896). Since
then little has been added to our knowledge of the species except that it lives
in mangrove areas in northern Australia and feeds on mudskippers
(Periopthahnus ) (Gow 1989; Cogger 1992). It is reported to swim along the
water line of rising or falling tides investigating crab holes in compacted to
relatively soft mud (Ehmann 1992). Six specimens were retrieved from the
stomachs of sharks that were netted within Darwin Harbour in the Frances
Bay region (Lyle & Timms 1987).

On 17 October 1992, we observed several H. daruhniensis from a dinghy on
Sadgrove's Creek, Darwin, where the species had been recorded previously (J.
Lylepers. comm.). Four hours of careful checking of the mangrove-lined streams
during low tide proved fruitless. At 15:00, however, as the tide began to rise, a
single H. darwiniensis was seen leaving the water and crawling onto a bank of
consolidated mud. It then moved into a nearby crab burrow to emerge some
minutes later from a neighbouring burrow. Another three individuals were seen
within 30 minutes of the first, but none in the subsequent two hours. These
observations, albeit limited, indicate that the species is diurnal, and that it
probably feeds out of the water.

Northern Territory Naturalist 14 (1993)

29

PLATE 5 The Port Darwin Sea Snake Hydrelnps darwiniensis from Sadgrove's
Creek (P.McGrath).

The first individual was captured and kept for 48 hours in a bucket of seawater
collected at the same time. The snake was extremely adept at moving on land, and was
placed in a pool in the upper tidal zone for photographing. It paused and drank for
about a minute from the pool which contained water that was fresh to taste and
presumably had accumulated from an overnight storm. Although sea snakes are
thought to obtain their fresh water requirements by removing salt from their bodies
with specialised "salt "glands near the tongue (Burns & Pickwell 1972), the closely-
related sea kraits are known to drink fresh water when it is available (Guinea 1991).

The terrestrial feeding habits of this species are unique among seasnakes, and
deserves further study.

References

Boulenger, G. A. (1896) Catalogue of the Snakes in the British Museum (Natural History).
Volume 3. Trustees of the British Museum, London.

Burns B, & Pickwell, G.V. (1972) Cephalic glands in sea snakes ( Pelamis , Hydrophis and
Laticauda). Copeia 3, 547-559.

30

Northern Territory Naturalist 14 <1993)

Cogger, H.G. (1992) Reptiles and Amphibians of Australia (5th edition). Reed International,
Sydney, 775 pp.

Ehmann, H. (1992) Encyclopedia of Australian Animals: Reptiles. Angus & Robertson, Pymble,
N.S.W.

Gow, G.F. (1989) Graeme Gow's Complete Guide to Australian Snakes. Angus and Robertson,
North Ryde, NSW.

Guinea, M.L. (1991) Rain water drinking by the sea krait Laticaudacolubrina. HerpetofaunaTl,
13-14.

Lyle, J.M. & Timms, G.J. (1987) Predation on aquatic snakes by sharks from Northern
Australia. Copeia 3, 802-803.

MICHAEL GUINEA*, PETER McGRATH and BARBARA LOVE, Faculty of Science,
Northern Territory University, PO Box 40146 Casuarina. NT.

^author for correspondence

Antiphonal Song in the Black Butcherbird
Cracticus quoyi in Darwin

Duetting, in which two birds sing simultaneously, is a common occurrence,
especially among medium sized tropical passerines (e.g. Diamond & Terborgh
1968; Diamond 1972). Antiphonal singing, in which two birds sing phrases
alternately to produce an integrated melody, is less commonly reported. Duetting
is well known in various Australian and New Guinea species of butcherbirds
(Macdonald 1973; Frith 1976; Coates 1990), but neither duetting nor antiphonal
song appear to have been recorded for the Black Butcherbird Cracticus quoyi. This
paper reports on the occurrence of antiphonal song for this species in the Darwin
region.

The Black Butcherbird is common in mangroves and remnant riparian forest in
the Darwin region. Between July 1988 and July 1993 I conducted an early morning
census of birds on about 18 days per month. The census followed a regular route
including residential streets, parkland, mangroves and riparian vegetation
bordering Rapid Creek. Two pairs of Black Butcherbirds, presumed to be the

Northern Territory Naturalist 14 (1993)

31

same resident individuals, were commonly recorded along sections of the route.
They were frequently seen and heard calling from mangroves, and less often
encountered in parkland and gardens, generally within about 50 metres of
mangroves.

Though less melodious than other butcherbirds, the Black Butcherbird has a
variety of rich calls. These show considerable geographic variation, at least in the
New Guinea region (Beehler et al. 1986; Coates 1990). In Darwin, the species is
particularly vocal at dawn when their loud calls may be heard up to a kilometre
away.

Antiphonal songs of Darwin birds were always composed of simple phrases in
two similar variants. The phrases were also commonly given by solitary
individuals. When singing antiphonally, two birds usually perched one to two
metres apart within the same tree, but on one occasion were in different trees,
about ten metres apart. One bird (Bird 1) was always perched higher, and began
each sequence of calls with a low-pitched, rather muted and slightly guttural (a)
double ("croo-croo") or (b) single ("ork") note. Bird 2 followed with louder,
higher-pitched, more melodious calls: (a) a disyllabic, upward-inflected "oo-wa",
or (b) a trisyllabic "oodle-oo" (Fig. f). In the latter trisyllabic call, the first and last
notes were of the same pitch; the second lower, and shorter in duration. Both
variants were delivered once or repeated several times, at the rate of three
sequences in five seconds.

Bird 1

Bird 1

croo croo

ork

Bird 2

r

Bird 2

~L_r

oo - wa

ooddle - oo

FIGURE 1. The tivo main song phrases of the Black Butcherbird

Antiphonal singing has only been observed clearly in the late dry season (Septem¬
ber to November) when birds are presumably preparing to breed. It may be
associated with establishing or re-affirming territories. Twice, 1 saw two pairs
engaged in bouts of aggressive pursuit interspersed with antiphonal song at a
territorial boundary. Since the species is sedentary and apparently maintains
territories throughout the year, however, it is possible that pairs engage in
antiphonal singing throughout the year. Further studies may clarify if this is so,
and if song varies within Australian populations.

32

Northern Territory Naturalist 7 4 (1993)

I thank Richard Noske and two anonymous referees for their comments on a draft
of this note.

References

Beeliler, B.M., Pratt, T.K., & Zimmerman, D.A. (1986) Birds of New Guinea. Princeton
University Press, Princeton, N.J.

Coates, B.J. 1990. The Birds of Papua Neic Guinea, Volume II. Dove Publications, Alderley,Qld.
Diamond, J.M. (1972) Further examples of duetting among southwest Pacific birds. Auk 89,
180-183.

Diamond, J.M. & Terborgh, J.VV. (1968) Dual singing bv New Guinea birds. Auk 85,62-82.
Frith, I I.J. (1976) Reader's Digest Complete Book of Australian Birds. Reader's Digest Services,
Sydney.

Macdonald, J.D. (1973) Birds of Australia: A Summary of Information. Reed, Sydney.

CHRISTOPHER HEALEY, Faculty of Arts, Northern Territory University, PO Box
40146 Casuarim. NT.

First Record of the Spectacled Monarch
Monarcha trivirgatus for the Northern Territory

The Spectacled Monarch Monarcha trivirgatus is found in coastal eastern
Australia, New Guinea, and many islands of eastern Indonesia, including
Timor (White & Bruce 1986). In far north Queensland it is sedentary, but
breeding populations in southern Queensland and northern New South Wales
migrate north during the winter (Blakers et at. 1984), probably mostly destined
for the Trans-Fly region of southern New Guinea (Beehler ct at. 1986). In
Australia it inhabits rainforests and occasionally mangroves (Blakers et at.
1984; Boles 1988), while in Timor it occurs in lowland evergreen and semi-
deciduous forests, as well as scrub (RN, pers. obs.).

During an avifauna! survey of mining leases on Groote Eylandt, we recorded
this species in a patch of coastal vine forest, c 24 km south of Alyangula, on 16
September 1991. Two birds were found: an adult and an immature, the latter
recognisable bv the lack of black on the face, the lores being whitish-grev, the
throat, dark grey (see Slater et at. 1986), and the base of the bill being pale
yellow. Both birds were present on the following day (RN), but the site was

Northern Territory Naturalist 14 (1993)

33

not re-visited until 5 January 1992, when an immature was sighted (GB). The
species was not found, however, during three visits to the site in late March
1992; nor was it ever recorded at the two other vine forest patches surveyed.

These records constitute the first of the Spectacled Monarch for the Northern
Territory, and the only known Australian records west of 141 °E (Cape York).
As the closest known locality to Groote Eylandt is Weipa on Cape York
(Blakers et al. 1984), it seems likely that the individuals concerned originated
from these populations, and were blown westward across the Gulf of
Carpentaria. The presence of an immature (albeit independent) bird in
September is curious given that Australian populations are purported to
breed from October to February (e.g. Boles 1988). Without information on the
duration of immature plumage in this species, however, the origin of these
birds must remain a mystery.

The above observations were made while conducting a bird survey for Groote
Eylandt Mining Company Pty Ltd, under the supervision of Mr Rob Savoury,
to whom we extend special thanks.

References

Beehler, B.M., Pratt, T.K. & Zimmerman, D.A. (1986). Birds of Neiv Guinea. Princeton
University Press, Princeton, New Jersey.

Blakers, M, Davies, S.J.J.F. & Reilly, P. (1984). The Atlas of Australian Birds. RAOU/
Melbourne University Press, Melbourne.

Boles, W.E. (1988) Robins and Flycatchers of Australia. National Photographic Index of
Australian Wildlife. Australian Museum/ Angus & Robertson, Sydney.

Slater, P., Slater, P. & Slater, R. (1986). The Slater Field Guide to Australian Birds. Rigby,
Sydney.

White, C.M.N & Bruce, M. (1986). Birds of Wallacea. British Ornithologists Club, London.

RICHARD NOSKE, Science Faculty, Northern Territory University, PO Box 40146
Casuarina, NT, 0811 and

GRAHAM BRENNAN, P.O. Box 33, Alyangula, Groote Eylandt, NT, 0885.

Northern Territory Naturalist 74 77 9 93)

CONTENTS

Effect of large herbivore exclusion on understorey biomass in three plant
communities on Cobourg Peninsula

D.M.J.S. BOWMAN 1

Assessment of Banteng grazing on coastal plains, Cobourg Peninsula:
an update on Bowman (7994)

W.J. PANTON 7

Seasonal patterns of wader populations in Darwin, Northern Territory,

1974-7987

K.S. SHURCLIFF and H.A.F. THOMPSON 10

Temporal dynamics of two tropical ant species, Iridomyrmex sanguineus

and Oecophytla smaragdina, in relation to temperature and humidity

A. M. LANE 79

SHORT NOTES

The use of small pebbles in the web retreat of a Comb-footed Spider,

Achacaranen sp. (Araneae: Theridiidae)

J. WEBBER 27

Observations of the Port Darwin Sea Snake Hydrelaps danciniensis

M. GUINEA, P. McGRATH and B. LOVE 28

Antiphonal song in the Black Butcherbird Cracticus quoyi in Darwin

C. HEALEY 30

First record of the Spectacled Monarch Monarcha trivirgatus for the

Northern Territory

R.A. NOSKE and’ G. BRENNAN

32