Diversity and Distributions, (Diversity Distrib.

) (2006) 12, 101–110

Blackwell Publishing, Ltd.

BIODIVERSITY Spatial dynamics of supercolonies of the

RESEARCH
invasive yellow crazy ant, Anoplolepis
gracilipes, on Christmas Island, Indian
Ocean
K. L. Abbott

School of Biological Sciences, Monash University, ABSTRACT

Melbourne VIC 3800, Australia
Correspondence: K. L. Abbott, School of
Biological Sciences, Victoria University of
Wellington, Wellington, New Zealand.
Tel.: +64 44635233 ext. 8135. Fax: +64 44635331.
E-mail: kirsti.abbott@vuw.ac.nz
Key to the management of invasive species is an understanding of the scope of an
invasion, the rate of proliferation and the rate at which invaded habitats become
degraded. This study examines the spatial dynamics of high-density supercolonies of
the invasive yellow crazy ant, Anoplolepis gracilipes, on Christmas Island, Indian
Ocean, and the associated impacts at their boundaries. Since the early 1990s, A. gracilipes
supercolonies have occupied over 30% of the 10,000 ha of rainforest on Christmas
Island. Thirty-four discrete high-density supercolonies formed between 1989 and
2003, ranging in size across nearly three orders of magnitude from 0.9 to 787 ha.
Supercolonies boundaries are diffuse, and ants were observed in low densities in
some cases up to 200 m from the main high-density supercolony. The 13 boundaries
examined were all dynamic over a 10–20 observation month period: nine bound-
aries expanded, and the maximum rate of spread was 0.5 m day−1. Across boundary
transition zones, between high-density supercolonies and intact rainforest, yellow
crazy ants reduced other ant species richness, occupied red crab burrows and killed
resident red crabs, which was the trigger for ‘invasional meltdown’ on Christmas
Island. The highly variable and unpredictable nature of A. gracilipes boundaries
poses a challenge for incorporation into a predictive framework, as well as for their
management.
Keywords
Anoplolepis gracilipes, biological invasions, invasive ants, oceanic island, rate of
spread, supercolony.

Several of the world’s worst invasive species are ants (Lowe

INTRODUCTION
In the 21st century, biological invasions by exotic plant and
animal species (whether through accidental or deliberate intro-
duction) rank amongst the greatest threats to native biodiversity
and ecosystem persistence (Vitousek et al., 1996; Mack et al.,
2000; Occhipinti-Ambrogi & Savini, 2003) and are economically
expensive (Leung et al., 2002; Perrings, 2002). While advances
are being made in the science of invasion ecology, much less
progress is being made in our ability to manage invasions
(Hulme, 2003). Key to their management is an understanding of
the scope of an invasion, the rate of spread and consequently the
rate at which invaded habitats become degraded. The area
occupied by alien species can determine the magnitude, cost and
likelihood of success of a management and subsequent control
program, while their rate of spread can dictate the urgency with
which such plans are implemented.
et al., 2000). Considering their increasing geographic range at
local, regional and international scales, invasive ant species
represent a significant conservation concern and threat to native
ecosystems, agricultural and tourism industries in both con-
tinental and island ecosystems ( Jahn & Beardsley, 1994; Holway
et al., 2002; O’Dowd et al., 2003). The Argentine ant, Linepi-
thema humile, has competitively displaced native ant species as it
spread in its introduced range (Erickson, 1971; Bond & Slingsby,
1984; Human & Gordon, 1996; Holway, 1998, 1999), and the red
imported fire ant, Solenopsis invicta, devastated native fauna as it
expanded its range across the southeastern United States (Allen
et al., 1995, 1998; Vinson, 1997). In monsoonal Australia, high
abundance of the big-headed ant, Pheidole megacephala, cor-
responded with a 42–85% decrease in the abundance of other
native invertebrates (Hoffmann et al., 1999), and the yellow
crazy ant, Anoplolepis gracilipes, displaced native spiders in

© 2006 The Author DOI: 10.1111/j.1366-9516.2006.00193.x 101

Journal compilation © 2006 Blackwell Publishing Ltd www.blackwellpublishing.com/ddi

K. L. Abbott
Hawaii (Gillespie & Reimer, 1993) and domestic animals in the
Seychelles (Haines & Haines, 1978b).
Anoplolepis gracilipes has been present on Christmas Island,
Indian Ocean, since the early 1900s (Donisthorpe, 1935), but its
distribution was unknown until populations exploded in the
early mid-1990s, and efforts were made to locate high-density
infestations. Defined as ‘supercolonies’, these regions of high-
density ant infestations progressively increased in number and
area to infest > 2500 ha of the 10,000 ha of rainforest by September
2002 (Green et al., 2004). Within supercolonies, the density of
yellow crazy ants on the forest floor alone is typically > 2000 ants
per m2 (Abbott, 2005); they lack distinct boundaries. Ants at low
densities can be observed hundreds of metres beyond main high-
density regions (cf. O’Dowd et al., 2003), as well as in isolated
nests around the island. In supercolonies, A. gracilipes rapidly
eliminated a keystone species: the red land crab, Gecarcoidea
natalis (Brachyura: Gecarcinidae), changing the composition
and structure of the forest (O’Dowd et al., 2003), resulting in
major ecosystem disruption and providing favourable condi-
tions for secondary invasions (Green et al., 2001). Supercolonies
of A. gracilipes are continuing to establish on Christmas Island,
concurrent with extremely high densities of honeydew-producing
scale insects that provide the ants with an abundant carbohydrate
source and possibly contribute to their success and continual
spread on the island (Abbott, 2004).
Studies in the Seychelles suggest that A. gracilipes can spread at
a rate of between 0.1 and 1.1 m day−1 (Haines & Haines, 1978a),
and impacts associated with high densities of yellow crazy ants
there include the disappearance of the Seychelles’ endemic skink,
sooty tern nesting failures on Bird Island in the Seychelles and
chicks killed by yellow crazy ants (Feare, 1999). On Christmas
Island, endemic red land crabs are killed within 48 h of exposure
to ants at extremely high densities (O’Dowd et al., 1999), the
abundance of the ground foraging emerald dove, Chalcophaps
indica, is significantly reduced in yellow crazy ant supercolonies,
and nesting success of the endemic Christmas Island thrush,
Turdus poliocephalus erythropleurus, might also be negatively
affected by A. gracilipes (Davis, 2002).
The majority of invasive ant colony boundary expansions,
including A. gracilipes, occur via a diffusion-like process (Holway,
1998; Hoffmann et al., 1999; Tsutsui & Suarez, 2003), while winged
queens flying to new nest sites and human-mediated introduction
of new propagules to more distant areas, illustrates the process of
middle- and long-distance jump dispersal (Suarez et al., 2001).
However, the relative contributions of dispersal by diffusion
(budding), middle-distance (flying queens) and long-distance
jump events (human-mediated dispersal of a viable propagule) to
the rate of spread for most invasive ants species is unclear (but see
Suarez et al., 2001 and Shigesada & Kawasaki, 1997). The tendency
for invasive ants to form unicolonial populations, and often super-
colonies, where ants cooperate over expansive areas of interconnected
nests (Holway et al., 1998; Giraud et al., 2002; O’Dowd et al., 2003;
Abbott, 2005) offers an opportunity to study local dispersal pro-
cesses such as spread by diffusion and middle-distance dispersal.
This study examines the distribution and spatial scale of high-
density A. gracilipes supercolony formation in an island-wide
102
Diversity and Distributions, 12, 101–110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
context, and documents the rate of movement of supercolony
boundaries, including both expansion and contraction. It
explores the nature of transition zones between high-density
supercolonies and intact rainforest that often extend hundreds of
metres beyond the main supercolony and present data that sug-
gest aerial dispersal of both queen and male A. gracilipes. It also
reports on impacts of expanding invasion boundaries: the effect
of A. gracilipes on species richness of the local ant fauna, red
crabs and crab burrows.
METHODS
Study system
This study was conducted on Christmas Island (105°40′ E,
10°30′ S), an isolated oceanic limestone island that lies 360 km
south of Java in the northeastern Indian Ocean. The island is
located in the humid tropics, and experiences a monsoonal cli-
mate where most of the annual 2000 mm rainfall occurs between
December and May (Falkland, 1986). Mean monthly air temper-
atures range between 20.8 °C min and 30.2 °C max (Australian
Bureau of Meteorology). Approximately 74% of the 135 km2
island is covered with natural vegetation. Three vegetation types
(I, II and III) dominate plateau (> 200 m a.s.l.) and terrace
(< 200 m a.s.l.) topography (Mitchell, 1975) and correspond
generally to geological substrate type (Barrie, 1967; Mitchell,
1975). Forest type I is described as tall closed forest on deep soil
with an average canopy height of 34 m; forest type II is closed
forest over shallower soils with an average canopy height of 28 m;
and forest type III is closed forest, mostly on a substrate of lime-
stone pinnacles and rocks with an average canopy height of 22 m.
Forest types II and III are drier and have a greater proportion of
deciduous species than forest type I. Mitchell (1975) provided a
map detailing the distribution of these forest types across the
island. Data for this study were collected between December
2000 and December 2002. Supercolonies were defined as areas
where the density of A. gracilipes foragers was sufficient to kill red
crabs, and this was a general operational definition on Christmas
Island for both managers and researchers.
Distribution of A. gracilipes supercolonies on
Christmas Island
The mapping of A. gracilipes supercolonies on Christmas Island
occurred in stages. Initially, researchers noticed an increase in the
abundance of A. gracilipes in 1996, and gathered information
regarding the distribution of high-density infestations. This was
done primarily through anecdotal evidence from Christmas
Island residents and recording their presence along 4WD and
accessible rainforest tracks (P.T. Green, pers. comm.). The dis-
tribution of A. gracilipes was verified with a formal island-wide
survey in April 2001, undertaken by Parks Australia and Monash
University. In that survey, 972 survey points distributed in a
regular grid pattern across the entire island were visited over a 3-
month period and assessed for yellow crazy ant activity using
card counts (see Green et al., 2004 for methods). These data
© 2006 The Author
 14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Spatial dynamics of yellow crazy ant

Table 1 Attributes and movement of 13 supercolony boundaries

on Christmas Island. Boundaries were divided by main substrate
type and observed for varying lengths of time. The mean transition
zone length is calculated from the length of the transition zones at
each time of observation (no. of observations in brackets). Boundary
number corresponds to numbers in Figs 2 and 5

Total Rate of Mean length

Months displacement movement transition zone
Boundary observed (m) (m year−1) (m ± SE)

Soil
1 20 120 72.0 46 ± 11 (8)
2 20 20 12.0 71 ± 10 (8)
3 11 90 98.4 114 ± 17 (5)
4 16 120 90.0 97 ± 8 (7)
Figure 1 Arrangement of transects perpendicular to Anoplolepis 5 20 90 54.0 56 ± 6 (6)
gracilipes supercolony boundary, extending into intact rainforest. 6 10 40 48.0 55 ± 9 (4)
The shaded region indicates the high-density supercolony where Scree slope
A. gracilipes kills red land crabs. The region extending outward to the 7 20 20 12.0 89 ± 11 (7)
right is the ‘transition zone’, where ant activity decreases from > 50 8 11 150 163.2 78 ± 17 (5)
ants per 30 s in the supercolony to zero in intact rainforest. The 9 15 −20 −15.6 98 ± 8 (6)
larger asterisk is the initial marker peg. The 2 × 6 m and 1 m2 10 11 30 32.4 53 ± 14 (4)
observation quadrats were centred on each 10-m interval point (X) 11 11 −30 −32.4 110 ± 15 (4)
where A. gracilipes activity was recorded. As boundaries expanded Steep soil and scree
(toward the right in figure) I extended the transect to reflect the 12 11 −10 −10.8 153 ± 19 (4)
change in boundary position. 13 20 −80 −48.0 105 ± 18 (8)
Mean 41.6 ± 18.9 53 ± 12 86.5 ± 8.5

formed the basis for subsequent detailed mapping of super-

colony boundaries between survey points by a relatively subjective
visual detection method. I used this information to produce a
map of the total area occupied by supercolonies of A. gracilipes
between 1996 and 2003. These areas did not include transition
zones outside the main supercolony (See Fig. 1 and below for
definition). Here I present the range in supercolony sizes and
the frequency of occurrence of each size class, as well as their dis-
tribution across local substrate types (deep soil, shallow soil with
scree and limestone).
Boundary movement, transition zones and flying
queens
The movement of 13 A. gracilipes supercolony boundaries was
documented over a period ranging between 10 and 20 months.
In November 2000, boundaries were chosen to include different
supercolonies that were relatively easy to access, and representing
a range of substrates and topographical situations, including
gentle soil slopes, scree slopes (small limestone pieces) and steep
limestone cliffs (Table 1). Observations were made every 3
months, weather permitting. The boundaries of supercolonies
were rarely sharp, and ant density near the edges of supercolonies
typically declined from very high to very low or zero over a dis-
tance of 10s or 100s of metres. I recorded yellow crazy ant
activity on the forest floor along a transect, perpendicular to this
decline in ant abundance, by counting the number of ants that
crossed a white laminated 10 × 10 cm card in 30 s (see Abbott,
2005 for methods). Cards were placed on the ground at 10-m
intervals along the linear transects until, beyond the supercolony,
ants were absent for at least 25 m, and into the supercolony until
ant activity was consistently ≥ 50 ants per 30 s for 30 m. The
supercolony boundary was defined as the point at which
A. gracilipes activity was ≥ 50 ants per 30 s, as this density of ants
appeared to kill red land crabs and was also used as an opera-
tional definition for management purposes. The extended zone
of declining ant activity beyond the supercolony boundary was
the ‘transition zone’. I placed a marker peg at the initial boundary
point as a reference point for future monitoring (Fig. 1). As
supercolony boundaries expanded and ant activity increased
beyond the initial supercolony, I extended the transect and
continued to record ant activity until A. gracilipes was absent for
at least 25 m. Given that the transition zone varied in width
between and within boundaries over time, the displacement of a
supercolony boundary was calculated as the total distance along
the transect between an initial and final ant activity count of 50
ants per 30 s.
In addition to A. gracilipes activity on the forest floor,
I recorded variables that allowed me to assess impacts at 10-m
intervals along transects. First, the number of other ant species
observed in a one minute visual search of a 1-m2 quadrat around
the observation point was recorded. Second, the number of red
crab burrows invaded by A. gracilipes, as well as intact red crab
burrows, and the number of dead crabs in a 2 × 6 m quadrat
around the observation point (Fig. 1). I looked for a correlation
between the A. gracilipes activity on the forest floor and other ant

© 2006 The Author 103

Diversity and Distributions, 12, 101– 110, Journal compilation © 2006 Blackwell Publishing Ltd
 14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
K. L. Abbott

Figure 2 The extent of Anoplolepis gracilipes

supercolony formation on Christmas Island
between 1996 and September 2002. Shaded
areas indicate densities of yellow crazy ants
that will kill red crabs and were defined as
supercolonies. Supercolony boundaries used
in this study are numbered (1–13) and
correspond with Table 1.

species richness (using raw data) and analysed the shape of the
relationship between the two variables using nonlinear regres-
sion on raw data. The relationship between ant activity and dead
crabs and invaded and intact crab burrows was analysed using
linear regression on raw data ( version 10).
The presence of winged queens and males in the canopy would
typically indicate mating flights, the timing and duration of dis-
persal events, or canopy nest sites. Aerial dispersal of winged
stages of A. gracilipes was monitored using one canopy malaise
trap at each of two sites on 19 October 2001, prior to the onset of
rain on 2 December 2002. Traps were checked weekly until 15
February 2002.

RESULTS
Figure 3 Range of supercolony areas (ha) on Christmas Island.
Distribution of Anoplolepis gracilipes supercolonies Supercolonies described here are those shaded in Figure 2.
on Christmas Island

Between 1996 and 2002, A. gracilipes supercolonies occupied a

total area of over 3000 ha of rainforest on Christmas Island
(> 30%; Fig. 2). Thirty-four supercolonies formed between 1989
(first high-density supercolony discovered) and September 2002,
ranging in size across nearly three orders of magnitude from
0.9 to 787 ha (Fig. 3). Supercolonies were irregular in shape and
often formed elongated ‘fingers’ that extended outward from a
core high-density region. Given their recent discovery and irregu-
larly timed formation, supercolonies of A. gracilipes, in combina-
tion with regions of lower ant density and isolated nests, have
formed a dynamic mosaic of ages and ant densities across the
entire island since 1989, but the lack of accurate distributional
data before 2001 prevents further analysis of these patterns.
Results from the island-wide survey demonstrated that the
dominant forest types (I, II and III) occurred over 32.4%, 41.0%
and 25.7% of the grid points surveyed across the island, respec-
tively. However, of all supercolony points in forest, 10% were in
forest type I, 54.7% were in type II and 35.3% were in type III.
This suggests that A. gracilipes formed supercolonies more fre-
quently than expected in forest types II and III, considering the
area of rainforest available for supercolony formation. Further-
more, 78.8% of all supercolonies were below 200 m a.s.l. (P.T.
Green for further unpublished analysis), most of which is con-
sidered ‘terrace’ or low elevation, drier, more open canopy forest,
but larger supercolonies spanned all three forest types. The forest

104 © 2006 The Author

Diversity and Distributions, 12, 101–110, Journal compilation © 2006 Blackwell Publishing Ltd

Figure 4 Mean (± SE) displacement (in relation to first
observation) of supercolony densities of Anoplolepis gracilipes along
transects between November 2000 and July 2002. Positive
displacement = boundary expansion, negative
displacement = boundary contraction. Horizontal dotted line
indicates zero net displacement.
types included a variety of substrate types, from soil, through to
soil interspersed with scree and limestone pinnacles, to areas
completely covered by limestone. The limestone slopes separat-
ing the terraces were often devoid of soil. I often observed nests
within pockets and cracks on limestone cliff faces, and foragers
were active on the surface of the rocks. There were nine discrete
supercolonies located in plateau rainforest on relatively deeper
soil and at higher elevation, but all three of the largest supercolonies
(448, 778 and 787 ha), and presumably the oldest, were primarily
located on the lower elevation terraces and were distributed
along the north coast, east coast and in the southwestern corner
Figure 5 Examples of supercolony boundary
expansion (a, left to right), contraction (b, right
to left) and amalgamation with a nearby
supercolony (c). In each example, ant activity
(ants across 10 × 10 cm card per 30 s) is shown
for stations spaced 10 m apart along a transect
running perpendicular to the supercolony
boundary, for both the initial observation
(
; November 2000 all sites) and last
observation (; July 2002 for boundary nos 5
and 13, October 2001 for boundary nos 8 and
12). The dotted line indicates yellow crazy ant
activity at densities sufficiently high to kill red
land crabs. The transition zone begins at the
point where ant activity drops below 50 ants per
30 s, and (d) is an example of a long transition
zone.
© 2006 The Author
Diversity and Distributions, 12, 101– 110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Spatial dynamics of yellow crazy ant
of the island, ranging in elevation from 20 m (terrace above the
water) to > 160 m a.s.l.
Boundary movement, transition zones and flying
queens
The rate of movement of supercolony boundaries was extremely
variable on Christmas Island. All 13 observed boundaries were
dynamic; nine out of 13 boundaries advanced over a period of
between 10 and 20 months of observation (Table 1). The highest
rate of expansion was 150 m in 11 months (0.50 m day−1), and
the slowest expanding boundary advanced 20 m in 20 months
(0.03 m day−1). The greatest contraction of a boundary was 80 m
in 20 months (0.13 m day−1), however, there was an overall net
expansion of supercolony boundaries during the observation
period (Fig. 4).
Transition zones spanned one, two and three substrate types
(e.g. soil, limestone scree and limestone pinnacles). There
appeared to be no pattern of movement associated with size of
the supercolony, nor any discernable pattern whereby ants
expanded or contracted into or out of particular substrate types.
However, 11 out of 13 boundaries moved across substrate
boundaries during the period of observation. Figure 5 illustrates
the spatial patterns of foraging ants along the transects, which
ran perpendicular to the invasion front. Over time, boundaries
expanded, contracted or fused with adjacent (often expanding)
supercolonies (Fig. 5a–c, respectively). Boundaries were rarely
sharp, and yellow crazy ant activity decreased gradually as obser-
vation points along the transects approached A. gracilipes-free
rainforest. The mean length of transition zones between super-
colonies of yellow crazy ants and undetectable yellow crazy ants
ranged from 46 ± 11 m to 153 ± 19 m (mean ± SE) (Fig. 5d).
Within transition zones, where densities of ants gradually
declined from within the main supercolony into intact rainforest,
105

K. L. Abbott
A. gracilipes coexisted with other ant species as well as red crabs,
G. natalis. As the activity of A. gracilipes increased to supercolony
densities at the initial boundary peg (i.e. as boundaries expanded),
other ant species richness declined. The relationship between A.
gracilipes activity and other species richness can be best described
with an exponential decay model, where yellow crazy ant activity
explains 46% of the variation within other ant species richness
(nonlinear curve estimation, R2 = 0.46). Where there was only
one other species of ant present in supercolonies, 81% of the
time it was Paratrechina minutula, which also appeared along the
transition zone into intact forest. Other ant species occasionally
encountered in supercolonies of A. gracilipes were Paratrechina
longicornis, a Pheidole sp., a Monomorium sp., Tetramorium spp.
and Tapinoma melanocephalum, all species distributed widely
throughout the tropics. Species most commonly found in transi-
tion zones, but not in supercolony densities, were P. longicornis,
Pheidole sp. and Brachyponera christmasi (possibly native).
Toward intact forest, where A. gracilipes activity was very low,
larger species such as Camponotus maculatus and Odontomachus
simillimus became more abundant.
As A. gracilipes activity increased at the initial boundary peg
(i.e. as boundaries expanded), the number of intact crab burrows
declined and the number of burrows used by yellow crazy ants
as nesting sites increased. After being occupied by ants for a
period of time, crab burrows lost their structural integrity and
collapsed, resulting in a total absence of crab burrows within
supercolonies that had been active for about 12 months or more.
Generally, the abundance of dead crabs increased and live crabs
decreased as A. gracilipes activity intensified. However, the abun-
dance of live and dead crabs depended upon their movement
through supercolony areas (which was higher during crab migra-
tion at the onset of rains usually in November/December) and
the initial density of red crabs in the area. Nonetheless, at the
initial marker peg, there was a significant relationship between
ant activity and the number of dead crabs observed at that point
on expanding boundaries only over time (F1,51 = 31.113, R2 = 0.379,
P < 0.001). As the density of crazy ants increased, they killed
resident red crabs.
Alate queens and males were caught in malaise traps > 15 m
above the ground at both ECR (three alate queens) and WBT
(57 queens and 37 males) at the onset of rains on 18 January and
25 January 2002, respectively.
DISCUSSION
The original entry point of A. gracilipes into Christmas Island
was the north-east of the island. Steep cliffs surround most of
the island, and there have only ever been two operational ports
(Flying Fish Cove and Waterfall Bay; see Fig. 2), both in the
north-east and both in use during the early part of last century
when yellow crazy ants are thought to have reached the island.
The first high-density supercolony was discovered approximately
halfway between these two sites in 1989 (P.T. Green, pers.
comm.). Since then, A. gracilipes has reached virtually every
corner of the island, but high-density supercolonies have formed
primarily in the western half of the island.
106
Diversity and Distributions, 12, 101–110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Dispersal of Anoplolepis gracilipes on Christmas Island
The spread, and therefore distribution, of A. gracilipes on Christ-
mas Island cannot be attributed to budding from one introduced
nest/colony. Even at the most rapid rate of spread by budding
observed in this study (163.2 m year−1), colonization of the entire
island by this method, possibly over the last 80 years, would be
improbable. The large physical gaps between supercolonies are
testimony to the fact that budding alone most likely did not pro-
duce the mosaic of infested areas. In addition to budding, there
are two other methods of dispersal that, combined, have more
likely aided the spread of A. gracilipes across Christmas Island,
and produced the pattern of high-density supercolonies seen
during the period between 1989 and 2002. First, and perhaps the
most important, is human-mediated transport, where prop-
agules were moved throughout the island during the early 1900s.
Ants could have hitched rides during exploration of the island
and when railways and roads were under construction (Gibson-
Hill, 1949), and this represents a likely explanation for coloniza-
tion of the southeastern and southwestern sections of Christmas
Island. Second, it is suspected that A. gracilipes queens disperse
by flying to uninfested areas to find a new colony of their own
(Abbott, 2004). It is aerial dispersal that would help explain
A. gracilipes’ establishment at locations over 2 km from high-
density infestations, and account for isolated nests in areas not
visited by people.
It is here that the distinction between the establishment and
spread of A. gracilipes and the formation of high-density super-
colonies becomes important. Establishment of A. gracilipes nests
represents an early stage of the invasion process (the engagement
of available propagules with a vector, transfer to recipient loca-
tion and establishment in recipient location), whereas the rapid
population explosion and formation of what we termed super-
colonies is the final stage of an invasion process — increase in
abundance to cause severe ecological and human impacts, and
this is the stage that is of interest in this study.
Supercolonies of A. gracilipes have formed in a temporally
asynchronous manner across the island; some supercolonies
increased in activity, while others declined (Abbott, 2004). The
increase in abundance occurred rapidly on Christmas Island, but
only after a considerable lag period (perhaps as long as 80–
90 years). The phenomena of lag times prior to population
explosions, whether rapid or gradual, is still poorly understood
(Simberloff & Gibbons, 2004), but are well known for some
invasions (Crooks & Soulé, 1999; Simberloff & Gibbons, 2004;
Memmott et al., 2005). They might be caused by the inherent
nature of population growth of a species, timing of environ-
mental factors related to improving ecological conditions for the
organism and a relative lack of fitness of the invader in its novel
environment (Crooks & Soulé, 1999). Physical conditions, medi-
ated by abiotic factors, have been shown to drive the increase
in abundance of pest insects in North America (Swetnam &
Lynch, 1993; Williams & Liebhold, 1995), where availability
of water and increased relative humidity are key factors in deter-
mining the distribution of the Argentine ant, L. humile (Holway,
1998; Walters & Mackay, 2003). Indeed, the environment on the
© 2006 The Author

Table 2 Reported rates of expansion and contraction of boundaries of other invasive ant species
Species Rate of movement (m year−1) Habitat /locality
Anoplolepis gracilipes Expansion 12–402 Seychelles, India, Christmas Island
Contraction 11–48
Linepithema humile Expansion 0–270 Worldwide
Pheidole megacephala Expansion 20–40 Northern Australia
Contraction 0–12
Solenopsis invicta Expansion 10–40 Texas
Wasmannia auropunctata Expansion 170–500 Santa Cruz Island, Galàpagos Archipelago
edges of Christmas Island might provide an indication of favour-
able conditions for the formation of supercolonies, and this is
mainly related to abiotic factors becoming favourable for their
scale insect mutualists (Abbott, 2004). However, there is a dearth
of research relating to mechanisms responsible for lag times,
especially into inherent characteristics of an invading organism,
and more is urgently required if predictive frameworks are a
priority for managers of invasions.
Boundary movement, transition zones and flying
queens
The expansion of supercolony boundaries on Christmas Island
represents diffusion by budding at a highly variable pace. This
rate of movement is within the range of Argentine ants, L. humile
(Suarez et al., 2001), but greater than the big-headed ant, P. meg-
acephala (Hoffmann et al., 1999), both species that disperse by
colony budding (See Table 2). Furthermore, for the supercolonies
studied here, only sections of an entire boundary moved (expanded
or contracted), and where and when it occurred was unpredict-
able. The net contraction of supercolony boundaries occurred
during sexual brood production (October–February; Abbott
2004), which is surprising considering that queens leave a
nest because of overcrowding (Hölldobler & Wilson, 1990).
Krushelnycky et al. (2003) suggest that dispersal (and hence
boundary expansion) of Argentine ants in Hawaii appears to occur
when worker production in the colony approaches its peak and
total ant abundance is high, but this did not seem to be the case
for A. gracilipes on Christmas Island. It is more likely that the
contraction of supercolony boundaries during wet periods was
observed due to increased rainfall, which depressed forager activity
on the surface of the ground, and on the activity cards used to
assess densities, and therefore boundaries. It does not indicate
that populations of A. gracilipes necessarily declined in these areas,
and it might not have represented a true boundary contraction,
but an absence of foragers on the ground. Conversely, reduced
rainfall during the long dry seasons experienced between 2000
and 2003 allowed for increased time outside the nest for workers
and queens, which might include travelling to new nest sites.
The variation in the measured rate of movement may also be
associated with extended transition zones. Boundaries of ant
colonies are traditionally defined as the point at which a specific
© 2006 The Author
Diversity and Distributions, 12, 101– 110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Spatial dynamics of yellow crazy ant
Source
Haines & Haines, 1978a; Rao et al., 1991;
This study
Cited in Suarez et al., 2001
Hoffmann et al. (1999)
Porter et al., 1988
Lubin, 1984, 1985
species fails to exist (or is no longer captured in pitfall traps).
However, by defining an A. gracilipes supercolony as the point at
which there were > 50 ants on the activity card per 30 s (and able
to kill red land crabs), that point will be subject to normal varia-
tion in forager activity at any point in time. The result is that the
detection of an absolute invasion front becomes difficult over
time (Haines & Haines, 1978a; cf. Argentine ant invasion fronts;
Erickson, 1971; Holway, 1998). The regions of interest in this
study were the high-density areas where A. gracilipes killed red
land crabs, and the detection and examination of this boundary
was considered key for managers. However, transition zones of
over 180 m represent an important gradient from extremely high
to low densities of A. gracilipes, within which a range of impacts
and interactions will occur. On one occasion, I observed over 100
dealate queens under a small log 50 m beyond the transition
zone, having not seen a worker since the edge of the transition
zone. Whether they flew (then shed their wings) or walked is
unknown.
Mating flights of A. gracilipes have not been previously docu-
mented (Rao et al., 1991; Haines et al., 1994), but Dammerman
(1929) suspected that mating flights could occur. The simultane-
ous capture of relatively high numbers of alate queens and males
on Christmas Island suggests that mating flights do occur at the
onset of rains, the main advantage being an increased rate of
spread. Alate queens were observed at and around fluorescent
light sources for up to three nights following the first rains of the
wet season in January 2001 and 2002 on Christmas Island, but
no dealates were observed at the same light sources. When a
sticky Tanglefoot™ barrier was placed around the trunks of the
Tahitian chestnut, Inocarpus fagifer, dealate queens were cap-
tured above the barrier, indicating they might have been nesting
in the canopy of the tree. If nests were abundant in the canopy,
then an alate queen produced there may choose to fly to find a
new colony, rather than nest in the same tree, where sites may
already be saturated.
In combination, colony budding and colony fission (flying
founder queens) are advantageous, and may provide adaptive
benefits, for several reasons. Colony budding results in colony
buds that are close together in space, often founded by already
mated queens. This can lead to polydomous colonies (having
multiple nest sites), increased kin recognition and cooperation,
fusion of colony boundaries and ultimately the formation of
107

K. L. Abbott
supercolonies (Hölldobler & Wilson, 1990; Passera, 1994; Bourke
& Franks, 1995). In contrast, colony fission results in fully in-
dependent daughter colonies (Pamilo, 1991), usually at greater
distances than those produced by colony budding. This affords
A. gracilipes greater dispersal powers, and the ability to disperse
by either method may even allow them to switch between modes
dependent upon colony status or environmental conditions.
Impacts of expanding Anoplolepis gracilipes
supercolonies
In this study, A. gracilipes displaced other ant species as their
activity increased and boundaries expanded, except for P. minutula
and P. longicornis, which were commonly found to coexist with
A. gracilipes in supercolony areas. This is in direct contrast to
patterns on Tokelau, where P. longicornis was never found
coexisting with A. gracilipes (Lester & Tavite 2004). I did not test
the underlying mechanism by which A. gracilipes dominated
those habitats, but the coexisting ant species were predominantly
small in size (with the exception of P. longicornis), and either able
to inhabit small interstitial spaces and run between crazy ant for-
agers or avoid excessive encounters with A. gracilipes by remaining
under organic matter on the forest floor. This observation is con-
sistent with the size–grain hypothesis, which predicts allometric
scaling of ant morphology (mass:leg length) that allows access to
different habitats within the same environment and subsequently,
coexistence (Kaspari & Weiser, 1999). Furthermore, foragers of
A. gracilipes monopolized food baits in supercolonies on Christ-
mas Island (Marr, 2003; Abbott, 2004), which indicated that
sheer abundance conferred them superior competitive and
exploitative abilities in those areas.
As an oceanic island, Christmas Island is naturally depauper-
ate of native ant species, or ant species in a similar functional
grouping (Andersen, 1995), which might bestow biotic resist-
ance and influence the rate of spread of A. gracilipes. It is unclear
whether there are any native ant species on the island, although
B. christmasi, among others, is a likely candidate (Donisthorpe,
1935). Therefore, encounters are occurring predominantly
between A. gracilipes and populations of other introduced ant
species, which would appear to be the case on many tropical
oceanic islands. The suite of coexisting ant species is becoming
increasingly dominated by introduced species, and the potential
for synergistic effects of these species is a concern for native
biodiversity. For example, the disappearance of red crabs from
supercolonies is due to direct aggression by A. gracilipes workers
(O’Dowd et al., 1999), but the protein source that dead crabs
supply to other introduced generalist ant species might be
important in their population dynamics on Christmas Island
too. However, this has not been tested.
Management implications
The dispersal capabilities of A. gracilipes, coupled with their
capacity to form extremely high-density supercolonies, indicate
that control and ongoing management of populations on Christ-
mas Island is more complex than simply monitoring existing
108
Diversity and Distributions, 12, 101–110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
infestations. Dispersal by budding will influence local move-
ment, where the size of each supercolony will dictate the magni-
tude of the control effort. However, the possibility of flights to
distant, uninfested areas by fertile queens means that continual
monitoring of uninfested rainforest is required to detect new
established nests and mitigate potential problems. Furthermore,
the spatial mosaic of varying sized A. gracilipes supercolonies
combined with the temporally asynchronous nature of suspected
‘boom and bust’ cycles of supercolonies, presents challenges for
managers. Thinking of high-density supercolonies as distinct
units for toxic bait control is useful, and enables priorities to be
set, where risk analyses can recommend a high likelihood of
success for control of particular supercolonies first. A reduction
in ant density in the highest density regions is crucial to the con-
tinued existence of the endemic red land crab, G. natalis, and the
preservation of Christmas Island rainforest. Waiting for a possible
natural decline of high-density populations would be detrimental
to the natural ecosystem.
Techniques for the control or eradication of invasive species
can take years to develop, and require ongoing commitment by
organizations. Control of over 2500 ha of A. gracilipes super-
colonies on Christmas Island was achieved in September 2002 by
the aerial distribution of toxic bait (Green et al., 2004), but since
then, numerous smaller supercolonies have formed in those
areas not targeted in 2002. Research into dispersal mechanisms of
alate queens, and what causes ‘booms and busts’ of A. gracilipes
supercolonies, will provide tools for predictive frameworks, and
equip managers with crucial knowledge for control practices.
Given immediate conservation threats and time constraints in
developing research and control programs to mitigate impacts
and understand invasion dynamics, it is not surprising that the
spatial dynamics, the tempo of invasion and its effects are key
issues in invasion biology today (Holway et al., 2002; Leung et al.,
2002; Krushelnycky et al., 2003). Knowledge of this kind is espe-
cially vital if management aims to prevent impacts rather than
respond to them.
ACKNOWLEDGEMENTS
I would like to thank Pete Green and Dennis O’Dowd for super-
vision and advice, the crazy ant survey team for finding ants in
the depths of Christmas Island rainforest, and Lori Lach, Stephen
Hartley and Jeremy Gibson for reviewing earlier versions of the
manuscript. Two anonymous reviewers also helped improve
the manuscript considerably. This work was funded by an
ARC-SPIRT grant awarded to DJO and PTG between Monash
University and Environment Australia.
REFERENCES
Abbott, K.L. (2004) Alien ant invasion on Christmas Island,
Indian Ocean: the role of ant-scale associations in the dynamics
of supercolonies of the yellow crazy ant, Anoplolepis gra-
cilipes. PhD Thesis, Biological Sciences. Monash University,
Melbourne.
Abbott, K.L. (2005) Supercolonies of the invasive yellow crazy
© 2006 The Author

ant, Anoplolepis gracilipes, on an oceanic island: Forager
patterns, density and biomass. Insectes Sociaux, 52, 266–273.
Allen, C.R., Lutz, R.S. & Demaris, S. (1995) Red imported fire ant
impacts on northern bobwhite populations. Ecological Appli-
cations, 5, 632–638.
Allen, C.R., Lutz. R.S. & Demaris, S. (1998) Ecological effects of
the invasive nonindigenous ant, Solenopsis invicta, on native
vertebrates: the wheels on the bus. Transactions of the North
American Wildlife and Natural Resources Conference, 63,
56–65.
Andersen, A.N. (1995) A classification of Australian ant com-
munities, based on functional groups which parallel plant
life-forms in relation to stress and disturbance. Journal of
Biogeography, 22, 15–29.
Barrie, J. (1967) The geology of Christmas Island. Bureau of
Mineral Resources of Australia Record 1967/37.
Bond, W. & Slingsby, P. (1984) Collapse of an ant-plant mutualism:
the Argentine ant (Iridomyrmex humilis) and myrmecochorous
Proteaceae. Ecology, 65, 1031–1037.
Bourke, A.F.G. & Franks, N.R. (1995) Social evolution in ants.
Princeton University Press, Princeton, New Jersey.
Crooks, J. & Soulé. M.E. (1999) Lag times in population
explosions of invasive species: causes and implications. Invasive
species and biodiversity management (eds O.T. Sandlund, P.J.
Schei & A. Viken), pp. 103–125. Kluwer Academic Publishers,
Dordrecht, the Netherlands.
Dammerman, K.N., ed. (1929) The agricultural zoology of the
Malay Archipelago. J.H. de Bussey, Ltd, Amsterdam.
Davis, N. (2002) The invasive yellow crazy ant (Anoplolepis
gracilipes) on Christmas Island, Indian Ocean: impacts on the
frugivorous bird fauna. BSc (Hons.) Thesis, School of Biological
Sciences. Monash University, Melbourne.
Donisthorpe, H. (1935) The ants of Christmas Island. Annual
Magazine of Natural History, 10, 629– 635.
Erickson, J.M. (1971) The displacement of native ant species by
the introduced Argentine ant Iridomyrmex humilis Mayr. Psyche,
78, 257–266.
Falkland, A. (1986) Christmas Island (Indian Ocean) water
resources — study in relation to a proposed development at
Waterfall. Hydrology and Water Resources Unit, Water Supply,
Sewerage and Stormwater Branch of the Transport and Works
Division for the Department of Territories, Canberra,
Australia.
Feare, C. (1999) Ants take over from rats on Bird Island, Seychelles.
Bird Conservation International, 9, 95–96.
Gibson-Hill, C.A. (1949) The early history of Christmas Island,
in the Indian Ocean. Journal of the Malayan Branch of the Royal
Asiatic Society, 22, 67–93.
Gillespie, R.G. & Reimer, N. (1993) The effect of alien predatory
ants (Hymenoptera: Formicidae) on Hawaiian endemic spiders
(Araneae: Tetragnathidae). Pacific Science, 47, 21–33.
Giraud, T., Pedersen, J.S. & Keller, L. (2002) Evolution of super-
colonies: the Argentine ants of Southern Europe. Proceedings of
the National Academy of Science (USA), 99, 6075–6079.
Green, P.T., O’Dowd, D.J. & Lake, P.S. (2001) From resistance to
meltdown: secondary invasion of an island rain forest. Interna-
© 2006 The Author
Diversity and Distributions, 12, 101– 110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Spatial dynamics of yellow crazy ant
tional Conference on Tropical Ecosystems. Oxford IBH, New
Delhi, Bangalore, India.
Green, P.T., Comport, S. & Slip, D. (2004) The management and
control of the invasive alien crazy ant (Anoplolepis gracilipes)
on Christmas Island, Indian Ocean: the aerial baiting campaign
September 2002. Unpublished final report to Environment
Australia and the Crazy Ant Steering Committee, Monash
University.
Haines, I.H. & Haines, J.B. (1978a) Colony structure, seasonality
and food requirements of the crazy ant, Anoplolepis longipes
(Jerd.), in the Seychelles. Ecological Entomology, 3, 109–118.
Haines, I.H. & Haines, J.B. (1978b) Pest status of the crazy ant,
Anoplolepis longipes (Jerdon) (Hymenoptera: Formicidae) in
the Seychelles. Bulletin of Entomological Research, 68, 627–638.
Haines, I.H., Haines, J.B. & Cherrett, J.M. (1994) The impact and
control of the crazy ant, Anoplolepis longipes (Jerd.), in the Sey-
chelles. Exotic ants: biology, impact, and control of introduced
species (ed. by D.F. Williams), pp. 206–218. Westview Press,
Boulder, Colorado.
Hoffmann, B.D., Andersen, A.N. & Hill, G.J.E. (1999) Impact of
an introduced ant on native rain forest invertebrates: Pheidole
megacephala in monsoonal Australia. Oecologia, 120, 595–604.
Hölldobler, B. & Wilson, E.O. (1990) The ants. Belknap Press,
Cambridge, Massachusetts.
Holway, D.A. (1998) Factors governing rate of invasion: a natural
experiment using Argentine ants. Oecologia, 115, 206–212.
Holway, D.A. (1999) Competitive mechanisms underlying
the displacement of native ants by the invasive Argentine ant.
Ecology, 80, 238–251.
Holway, D.A., Suarez, A.V. & Case, T.J. (1998) Loss of intra-
specific aggression in the success of a widespread invasive
social insect. Science, 282, 949–952.
Holway, D., Lach, L., Suarez, A.V., Tsutsui, N.D. & Case, T.J.
(2002) The causes and consequences of ant invasions. Annual
Review of Ecology and Systematics, 33, 181–233.
Hulme, P.E. (2003) Winning the science battles, but losing the
conservation war? Oryx, 37, 178–193.
Human, K.G. & Gordon, D.M. (1996) Exploitation and interference
competition between the invasive Argentine ant, Linepithema
humile, and native ant species. Oecologia, 105, 405–412.
Jahn, G.C. & Beardsley, J.W. (1994) Big-headed ants, Pheidole
megacephala: interference with the biological control of gray
pineapple mealybugs. Exotic ants biology, impact, and control of
introduced species (ed. by D.F. Williams), pp. 199–205. West-
view Press, Boulder, Colorado.
Kaspari, M. & Weiser, M.D. (1999) The size-grain hypothesis
and interspecific scaling in ants. Functional Ecology, 13, 530–
538.
Krushelnycky, P.D., Loope, L.L. & Joe, S.M. (2003) Limiting
spread of a unicolonial invasive insect and characterization of
seasonal patterns of range expansion. Biological Invasions, 6,
47–57.
Lester, P.J. & Tavite, A. (2004) Long-legged ants (Anoplolepis gra-
cilipes) have invaded Tokelau, changing the composition and
dynamics of ant and invertebrate communities. Pacific Science,
58, 391–401.
109

K. L. Abbott
Leung, B., Lodge, D.M., Finnoff, D., Shogren, J.F., Lewis, M.A. &
Lamberti, G. (2002) An ounce of prevention or a pound of
cure: bioeconomic risk analysis of invasive species. Proceedings
of the Royal Society of London. Series B, Biological Sciences, 269,
2407–2413.
Lowe, S., Browne, M. & Boudjelas, S. (2000) 100 of the world’s
worst invasive alien species. Aliens, 12, S1–S12.
Lubin, Y.D. (1984) Changes in the native fauna of the Galápagos
Islands following invasion by the little red fire ant, Wasmannia
auropunctata. Biological Journal of the Linnean Society, 21,
229–242.
Lubin, Y.D. (1985) Studies of the little red fire ant, Wasmannia
auropunctata, in a Nino year. El Nino en las Islas Galápagos: el
evento 1982–1983 (ed. by G. Robinson & E.M. del Pino),
pp. 473 –493. Charles Darwin Foundation, Quito, Ecuador.
Mack, R.N., Simberloff, D., Lonsdale, W.M., Evans, H., Clout, M.
& Bazzaz, F.A. (2000) Biotic invasions: causes, epidemiology,
global consequences, and control. Ecological Applications, 10,
689–710.
Marr, R. (2003) Toxic baiting of the yellow crazy ant, Anoplolepis
Gracilipes, on Christmas Island: impacts on non-target litter
invertebrates. BSc (Hons) Thesis, School of Biological Science.
Monash University, Melbourne.
Memmott, J., Craze, P.G., Harman, H.M., Syrett, P. & Fowler, S.V.
(2005) The effect of propagule size on the invasion of an alien
insect. Journal of Animal Ecology, 74, 50–62.
Mitchell, B.A. (1975) The forest flora of Christmas Island. Com-
monwealth Forestry Review, 53, 19–29.
O’Dowd, D.J., Green, P.T. & Lake, P.S. (1999) Status, impact and
recommendations for research and management of exotic invasive
ants in Christmas Island National Park. Report to Environment
Australia, Monash University, Melbourne.
O’Dowd, D.J., Green, P.T. & Lake, P.S. (2003) Invasional ‘melt-
down’ on an oceanic island. Ecology Letters, 6, 812–817.
Occhipinti-Ambrogi, A. & Savini, D. (2003) Biological invasions
as a component of global change in stressed marine ecosystems.
Marine Pollution Bulletin, 46, 542–551.
Pamilo, P. (1991) Evolution of colony characteristics in social
insects. I. Sex allocation. American Naturalist, 137, 83–
107.
Passera, L. (1994) Characteristics of tramp species. Exotic
110
Diversity and Distributions, 12, 101–110, Journal compilation © 2006 Blackwell Publishing Ltd
14724642, 2006, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1366-9516.2006.00193.x by Cochrane Germany, Wiley Online Library on [11/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ants: biology, impact, and control of introduced species (ed. by
D.F. Williams), pp. 23–43. Westview Press, Boulder, Colorado.
Perrings, C. (2002) Biological invasions in aquatic systems: The
economic problem. Bulletin of Marine Science, 70, 541–552.
Porter, S.D., Van Eimeren, B. & Gilbert, L.E. (1988) Invasion of red
imported fire ants (Hymenoptera: Formicidae): microgeography
of competitive replacement. Annals of the Entomological Society
of America, 81, 913–918.
Rao, N.S., Veeresh, G.K. & Viraktamath, C.A. (1991) Dispersal and
spread of crazy ant Anoplolepis longipes ( Jerdon) (Hymenoptera:
Formicidae). Environmental Ecology, 9, 682–686.
Shigesada, N. & Kawasaki, K. (1997) Biological invasion: theory
and practice, pp. 79–102. Oxford University Press, New York.
Simberloff, D. & Gibbons, L. (2004) Now you see them, now you
don’t! — population crashes of established introduced species.
Biological Invasions, 6, 161–172.
Suarez, A.V., Holway, D.A. & Case, T.J. (2001) Patterns of spread
in biological invasions dominated by long-distance jump
dispersal: insights from Argentine ants. Proceedings of the
National Academy of Sciences, USA, 98, 1095–1100.
Swetnam, T.W. & Lynch, A.M. (1993) Multicentury, regional-scale
patterns of western spruce budworm outbreaks. Ecological
Monographs, 63, 399–424.
Tsutsui, N.D. & Suarez, A.V. (2003) The colony structure and popu-
lation biology of invasive ants. Conservation Biology, 17, 48–58.
Vinson, S.B. (1997) Invasion of the red imported fire ant
(Hymenoptera: Formicidae): spread, biology, and impact.
American Entomologist, 43, 23–39.
Vitousek, P.M., D’Antonio, C.M., Loope, L.L. & Westbrooks, R.
(1996) Biological invasions as global environmental change.
American Scientist, 84, 468–478.
Walters, A.C. & Mackay, D.A. (2003) An experimental study of
the relative humidity preference and survival of the Argentine
ant, Linepithema humile (Hymenoptera, Formicidae): com-
parisons with a native Iridomyrmex species in South Australia.
Insectes Sociaux, 50, 355–360.
Williams, D.W. & Liebhold, A.M. (1995) Forest defoliators and
climatic change: potential changes in spatial distribution of
outbreaks of western spruce budworm (Lepidoptera: Tortricidae)
and gypsy moth (Lepidoptera: Lymantriidae). Environmental
Entomology, 24, 1–9.
© 2006 The Author