HOUSEHOLD AND STRUCTURAL INSECTS

Relative Attractiveness of Baits to Paratrechina longicornis

(Hymenoptera: Formicidae)
MARGARET C. STANLEY1 AND WAYNE A. ROBINSON2

J. Econ. Entomol. 100(2): 509Ð516 (2007)

ABSTRACT Exotic ant incursions are becoming more frequent around the globe, and management
with toxic baits is a suitable strategy for most species. Crazy ants, (Latreille) (Hymenoptera: For-
micidae), however, are notoriously difÞcult to attract to commercial baits, which are generally tailored
to the preferences of Þre ants. We tested P. longicornis preferences for various food types and
commercial ant baits. Baits trialed were commercially available products Amdro, Maxforce, Xstinguish
(nontoxic monitoring version), Presto, and tuna (in spring water), sugar water (25%), boric acid (1% in
25% sugar water), and deionized water. Tuna and Xstinguish, along with sugar water and sugar water ⫹

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

boric acid, were the most attractive baits to P. longicornis foragers. The granular baits (Maxforce,
Amdro, and Presto) were not as attractive to P. longicornis foragers. A decrease in temperature from
summer (30⬚C) to autumn (23⬚C) trials did not seem to affect the food preferences of P. longicornis.
Although P. longicornis recruitment was substantially lower during trials where there was concurrent
high native ant abundance and diversity, P. longicornis still recruited to preferred baits in numbers
higher than any other species. Given that tuna is impractical for management programs, the effec-
tiveness of boric acid, sweet liquid baits in eliminating P. longicornis colonies should be compared with
that of the toxic version of Xstinguish. If both are effective at eliminating colonies, we recommend
sweet liquid baits containing boric acid be used for small-scale incursions (one or two nests), but a
more practicable solid bait, such as Xstinguish, be used for larger scale incursions (numerous nests).

KEY WORDS Paratrechina longicornis, crazy ants, bait preference, ant management

Exotic ant invasions can be ecologically destructive other Paratrechina species (Stanley 2004). There is
and have major economic and social impacts (Holway little research on the food and bait preferences of P.
et al. 2002). A frequent management strategy is to longicornis, although indoor surveys in Malaysian res-
attract the ants to toxic baits, but bait acceptance is idential areas found honey (80% of workers) was
crucial to their success. Foraging ants must be at- strongly preferred by P. longicornis over peanut butter
tracted to the bait, must feed on the bait sufÞciently, (20% workers) (Lee 2002). Lee and Tan (2004) re-
and must carry it back to the nest and share it with ported that baiting is seldom effective, particularly
other members of the colony (Davis and van Schagen with paste and granular commercial formulations,
1993, Klotz and Williams 1996, Collins and Callcott against P. longicornis in Singapore and Malaysia and
1998, Lee 2000). Ant preferences for different food recommended sugar-based, liquid, or gel formulations
types (e.g., proteins, carbohydrates, and lipids), dif- for control of P. longicornis (Lee 2002). Tuna (in oil)
ferent-sized particles, and seasonal variation in these baits used in Biosphere 2 (in which P. longicornis was
preferences will determine how appropriate toxic the dominant ant) were consistently more attractive
baits are for use against particular ant species. to P. longicornis than the pecan cookie baits (primarily
Paratrechina longicornis (Latreille) (Hymenoptera: carbohydrate) put out at the same time (Wetterer
Formicidae) is notoriously difÞcult to control with et al. 1999; J. Wetterer, personal communication).
bait (Hedges 1996a, 1996b; Lee 2002). Hedges (1996a) Observations during incursions in New Zealand have
reported P. longicornis would not feed for long enough revealed that P. longicornis prefer sweet baits over
on commercial baits to ensure effective control. How- protein baits but that they recruit well to the protein-
ever, commercial ant baits are usually tailored to the based Xstinguish bait (T. Ashcroft, personal commu-
preferences of red imported Þre ant, (Solenopsis in- nication). Other species of Paratrechina have been
victa Buren), and it is likely that oil-based baits are not observed to recruit well to protein and sugar water
preferred by P. longicornis, given observations by baits, but they largely ignore Maxforce and Amdro
other researchers (see below) and the preferences of granules (Harris et al. 2002, Krushelnycky and Lester
2003). However, no formal testing of bait attractive-
1Landcare Research, Private Bag 92170, Auckland, New Zealand.
ness has been carried out for P. longicornis.
2University of the Sunshine Coast, Maroochydore DC Qld 4558, This research trialed the relative attractiveness of
Australia. various food types and baits to P. longicornis, with the

0022-0493/07/0509Ð0516$04.00/0 䉷 2007 Entomological Society of America

510 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 2

Table 1. Baits used in trials for relative attractiveness of food types to P. longicornis

Primary nutrient
Bait Quantity
Bait class of bait Manufacturer
formulation (g)
matrix
Amdro (⫹0.73% hydramethylnon) Lipid Granules 5 Ambrands (BASF Corporation), Australia
Boric acid ⫹ sugar water (1% boric acid) Carbohydrate Liquid 3
Deionized water (no toxin) Water Liquid 3
Maxforce (⫹1% hydramethylnon) Protein Granules 5 Bayer Environmental Science, USA
Presto (⫹0.01% Þpronil) Protein Granules 5 BASF Australia, Australia
Sugar water (25% sucrose) (no toxin) Carbohydrate Liquid 3
Tuna (no toxin) Protein Solid 5 South Australian Fisheries Company Limited
(SAFCOL)
Xstinguish (nontoxic version) Protein and Paste 5 Bait Technology Ltd., New Zealand
carbohydrate

Liquid baits were a standard of three grams equating to one soaked cotton dental swab (38 by 10 mm). In trials 1, 3, 4, and 5 tuna was used
instead of Presto (Þsh meal matrix) due to the unavailability of this product. Boric acid is analytical reagent boric acid (99.5%). For more
information on product details, see Stanley (2004). Formulations of commercially available baits are not accessible due to commercial sensitivity.

ultimate aim of determining its food preferences and tigated how bait attractiveness and recruitment of P.

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

the appropriateness of various commercial ant baits longicornis might differ at a site where there were a
for P. longicornis management. The attractiveness of variety of behaviorally dominant native ant species
seven bait types to P. longicornis was tested in summer present (site 2). Trial 4 investigated whether P. lon-
and autumn Þeld trials in southeastern Queensland, gicornis might avoid interacting with native ant species
Australia. Recruitment of P. longicornis workers to the by foraging at night. Because few ants (of all species)
baits in the presence of other ant species also was foraged on baits at night in trial 4, a smaller trial (trial
investigated. 5) was carried out the next day to investigate whether
low numbers were a result of the time of day effect or
Materials and Methods whether ant density in the area had decreased in the
3 wk since trial 3.
The research was undertaken at KingÞsher Bay Re- Baits trialed were the commercially available prod-
sort and Village (KBRV), Fraser Island, Queensland, ucts Amdro, Maxforce, Xstinguish (nontoxic monitor-
Australia. Fraser Island is a vegetation-covered sand ing version), and Presto as well as tuna (in spring
island located ⬇290 km north of Brisbane (27⬚ 37⬘ 27 water), sugar water (25% sucrose), boric acid (1% in
S, 153⬚ 05⬘ 32 E). KingÞsher Bay is on the western side 25% sugar [sucrose] water), and deionized water (Ta-
of the Fraser Island, 15 km east of Hervey Bay on ble 1). This ensured all major food types (e.g., pro-
mainland Australia. KBRV is a minimal-impact ecore- teins, carbohydrates, and lipids) were available. Seven
sort set within native vegetation (canopy trees are baits were used in each trial, and tuna was used when
primarily native pines and eucalypts). The climate is
Presto was unavailable (trials 1, 3, 4, and 5) to ensure
subtropical.
a Þsh-based bait type was included in the trials.
The boundaries of the P. longicornis infestation at
Daytime trials were timed to coincide with after-
KBRV were identiÞed in the Þrst weeks of March 2005
noon peak foraging activities (W.A.R., unpublished
by using observations (aggression assays) and honey
data), and ambient temperatures were recorded dur-
baits. There is only one large supercolony spread over
several hectares and included areas of high P. longi- ing each sampling round. During each trial, replicate
cornis numerical dominance (low abundance of other bait stations (trials 1, 3, and 4, n ⫽ 32; trial 2, n ⫽ 31;
ant species) as well as areas where high numbers of P. and trial 5, n ⫽ 11) were positioned at least 10 m apart
longicornis frequently co-occurred with native ants. within the trial site. At each bait station, each of the
Site 1 was identiÞed as an area where complete nu- bait types were placed on one of seven white petri
merical dominance of P. longicornis was expected and dishes positioned in random order in a circular or oval
was used for a summer and an autumn trial. Site 2 was pattern with at least 30 cm between dishes. The po-
a site where P. longicornis co-occurred with native ants sition order of baits at each station was determined
and was used in an autumn trial. Site 2 was ⬇50 m using the PLAN procedure in SAS (SAS Institute
down a heavily vegetated incline from site 1, but it 1992). Recruitment (number of workers in petri dish)
offered completely different environment. Although to each of the seven baits was recorded at each bait
site 2 was completely shaded with trees up to 30 m in station at least three times (three sampling rounds).
height, site 1 had few tall trees, much drier leaf litter Four sets of observations for each bait station (four
and provided only partial shade. sampling rounds) were made during trials 1, 3, and 5.
Overall, Þve separate trials were performed. Trials Each sampling round took ⬇30 min, although the
1 (summer) and 2 (autumn) tested the attractiveness night trial (trial 4) took longer to complete and was
of the baits to P. longicornis at a site where P. longi- reduced to three sampling rounds. A storm also re-
cornis was dominant (site 1). Trials 3 and 4 were duced the number of sampling rounds completed for
performed in autumn, but 3 wk apart, and they inves- trials 2 to three sampling rounds.
April 2007 STANLEY AND ROBINSON: ATTRACTIVENESS OF BAITS TO P. longicornis 511

Fig. 1. Mean abundance of P. longicornis workers on baits during bait preference trial 1: summer, P. longicornis numerically
dominant (n ⫽ 32 bait stations). Standard errors omitted to retain clarity. Means with the same letter at each time are not signiÞcantly
different according to TukeyÕs honestly signiÞcant difference multiple comparisons procedure.

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

Recruitment was recorded by estimating the num- Results
ber of ants on each bait. During trial 1, a digital image
Trial 1: Summer, P. longicornis Numerically Dom-
was taken of the petri dish for each bait type at every inant. At site 1 where P. longicornis dominates the ant
Þfth bait station. The digital images were enlarged, and community, P. longicornis foragers were present at all
the actual abundance of ants on each bait was deter- bait stations. With the exception of two Amdro baits
mined. A regression analysis of actual versus estimated and one water bait, P. longicornis foragers were ob-
numbers was performed, and the slope of the line used served on all baited petri dishes (n ⫽ 224) at least once
to correct the estimated abundance before analysis. during the 2-h period. Concordance between bait
For trials 2, 3, 4 and 5, counts were calibrated in situ preferences was extremely high and consistent across
against a standard score sheet that contained a range the different sampling rounds (␹2 ⫽ 169.8, df ⫽ 6, P ⬍
of the digital images and counts from the Þrst trial. The 0.0001).
same observer (W.A.R.) made all abundance esti- There was a signiÞcant difference in the number of
mates across all trials. When ant species other than P. P. longicornis foragers across the four observation
longicornis was present on a bait, and the species could times (F ⫽ 30.99; df ⫽ 3, 558; P ⬍ 0.0001) and across
not be identiÞed in situ (by W.A.R.), workers were the seven bait types (F ⫽ 471.2; df ⫽ 6, 186; P ⬍
collected and identiÞed using the Fraser Island 0.0001), and the two effects were not independent
Voucher collection held in the University of the Sun- (Ftime ⫻ bait ⫽ 2.26; df ⫽ 18, 558; P ⬍ 0.005). The
shine Coast invertebrate bioassessment program. posthoc analyses found consistent trends in prefer-
Using relative abundance of P. longicornis as an indi- ence across time (Fig. 1). There was high forager
cation of preference, the CochranÐMantelÐHaenzel (C- recruitment to both the Xstinguish and tuna baits, and
M-H) statistic was used to test for concordance in the these baits were preferred over the sugar water baits
order of preference of bait type across the replicate (sugar water and sugar water ⫹ boric acid) through
bait stations. The C-M-H test is not inßuenced by large the Þrst three sampling rounds; however, sugar water
numbers and uses rank order within each bait station; was not statistically different to the Xstinguish and
hence, it automatically partitions differences between tuna baits by the last sampling round (Fig. 1). Al-
though Maxforce had slightly higher worker abun-
bait stations. A repeated measures analysis of variance
dance than Amdro, the granular baits along with the
(ANOVA) (bait stations treated as blocks) was used
water had very low, almost nil, forager recruitment
to determine whether abundance changed across (Fig. 1). Increases in forager recruitment to the sugar
time, across baits, and whether there was an interac- water baits were maintained throughout the trial, but
tion of time ⫻ bait. SigniÞcant effects were deter- recruitment leveled off or decreased at the tuna and
mined using TukeyÕs multiple comparisons procedure. Xstinguish baits (Fig. 1), possibly because workers
Ant abundance was transformed using Y⬘ ⫽ log could remove large chunks of these baits from the
10(Y ⫹ 1) to reduce the effects of skewness in abun- petri dishes.
dance. Furthermore, because of the potential for lack Trial 2: Autumn, P. longicornis Numerically Dom-
of sphericity in the repeated measures (because abun- inant. Temperatures averaged 23⬚C during this trial,
dance can be high on some baits and expected to vary compared with 28 Ð31⬚C during trial 1 in summer. P.
through time), the effect of time and the interaction longicornis foragers were present at 181 of 217 baited
effect was adjusted using GreenhouseÐGeisser cor- petri dishes (n ⫽ 31 bait stations), but they were not
rection. observed at all on 45% of the water baits, and 32, 29,
512 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 2

Fig. 2. Mean abundance of P. longicornis workers on baits during bait preference trial 2 (n ⫽ 31 bait stations). Standard
errors omitted to retain clarity. Means with the same letter at each time are not signiÞcantly different according to TukeyÕs
honestly signiÞcant difference multiple comparisons procedure.

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

and 23% of the Presto, Amdro, and Maxforce baits, Trial 3: Autumn, P. longicornis Co-Occurs with
respectively. P. longicornis foragers were present on all High Abundance of Native Ant Species. The Xstin-
but one of the other three baits (Xstinguish, sugar guish used in the Þrst 24 bait stations during trial 3 had
water, sugar water ⫹ boric acid). There was strong been left over from a previous trial, and although
concordance among bait preferences, which was con- sealed, the product was probably stale (evidenced by
sistent across the different sampling rounds (␹2 ⫽ the attendance of ßies for at least the Þrst eight of
155.8, df ⫽ 6, P ⬍ 0.0001). There was a signiÞcant those stations). After the Þrst count (40 min), fresh
difference in the number of P. longicornis foragers bait was applied at these 24 stations. The statistical
across the three observation times (F ⫽ 10.16; df ⫽ 2, analyses presented here are therefore from the Þnal
360; P ⬍ 0.0001) and across the seven bait types (F ⫽ eight stations only. Each sampling round in this trial
453.3; df ⫽ 6, 180; P ⬍ 0.0001), and the two effects were took ⬇40 min (rather than the usual 30 min) because
not independent (Ftime ⫻ bait ⫽ 4.86; df ⫽ 12, 360; P ⬍ of the time taken to collect other ant species feeding
0.0001). Posthoc analyses found consistent trends in on the baits. P. longicornis was the most common of the
preference across time (Fig. 2); Xstinguish attracted 19 ant species that visited the baits, and it was observed
more foragers than any other bait, whereas sugar wa- visiting between 10 and 29 of the 32 petri dishes for
ter and sugar water ⫹ boric acid were also more each bait type.
attractive to foragers than the granular baits and water The concordance among bait preferences by P. lon-
(Fig. 2). Presto was no more attractive to foragers than gicornis was again high across all sampling rounds
deionized water (Fig. 2). Overall, forager abundance (␹2 ⫽ 17.9, df ⫽ 6, P ⬍ 0.01). P. longicornis decreased
was comparable with that in trial 1, even though trial in abundance on tuna baits but increased on boric
2 did not go the full four sampling rounds. Hence, the acid, sugar water, and Xstinguish during the four sam-
change in season (this trial being 6 wk later in the pling rounds (Fig. 3).
season than trial 1 and ambient temperature ⬇6⬚C There was no signiÞcant difference in the number
lower) did not seem to affect the numbers of active of P. longicornis foragers across the four observation
foragers. times (F ⫽ 2.53; df ⫽ 3, 102; P ⫽ 0.1), but there was

Fig. 3. Mean abundance of P. longicornis workers on baits during bait preference trial 3 (n ⫽ 8 bait stations). Standard
errors omitted to retain clarity. Means with the same letter at each time are not signiÞcantly different according to TukeyÕs
honestly signiÞcant difference multiple comparisons procedure.
April 2007 STANLEY AND ROBINSON: ATTRACTIVENESS OF BAITS TO P. longicornis 513

Fig. 4. Mean abundance of P. longicornis workers on baits during bait preference trial 5 (n ⫽ 11 bait stations). Standard
errors omitted to retain clarity. Means with the same letter at each time are not signiÞcantly different according to TukeyÕs
honestly signiÞcant difference multiple comparisons procedure.

a difference across the seven bait types (F ⫽ 16.4; dance in preferences (␹2 ⫽ 46.4, df ⫽ 6, P ⬍ 0.0001).

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

df ⫽ 6, 34; P ⬍ 0.0001), and there was no interaction Xstinguish was the most attractive bait and water the
(Ftime ⫻ bait ⫽ 1.03; df ⫽ 18, 102; P ⫽ 0.43). The posthoc least attractive across all species.
analyses found that the granular baits were not pre- Trial 5: Late Autumn Trial, P. longicornis Co-Oc-
ferred by P. longicornis foragers, but numbers were not curred with Low Abundance of Native Ant Species. Be-
as high on Xstinguish as in the previous two trials cause ant abundance was so low during the night trial
relative to the sugar or tuna baits (Fig. 3). On the sugar (trial 4), we set out 11 bait stations at the same site the
water, sugar water ⫹ boric acid, and tuna baits, P. next day to see whether ant abundance during the day
longicornis was the most abundant ant, averaging 60 Ð was comparable with the previous trial conducted
80% of all individuals. P. longicornis made up ⬇40% of there (trial 3). The air temperature was 24⬚C through-
all ants recorded on Xstinguish but ⬍10% on Amdro out this trial, the same as experienced throughout trial
and Maxforce. The number of records (presence/ 3 (3 wk earlier). However, the numbers of P. longi-
absence) of P. longicornis relative to all other ant cornis foraging during trial 5 were about half those
species was in similar proportions to the abundance. observed during trial 3 (Fig. 4). Two Xstinguish and
Trial 4: Late Autumn Night Trial, P. longicornis two tuna baits were consumed by a lace monitor,
Co-Occurred with Low Abundance of Native Ant Spe- Varanus varius, between the second and third counts
cies. Ant abundance during the night trial (trial 4) was of this trial, and the numbers for these baits were
considerably lower than in trials 1Ð3. Temperatures affected; hence, counts 3 and 4 from all baits on these
averaged 17Ð19⬚C during this trial, compared with bait stations were removed from the analysis. Con-
21Ð24⬚C during the diurnal autumn trials at the same cordance among bait preferences was again very
site. Crematogaster cornigera Forel, Aphaenogaster lon- strong for P. longicornis (␹2 ⫽ 36.5, df ⫽ 6, P ⬍ 0.0001).
giceps (Smith), and P. longicornis were the most com- There was a signiÞcant difference in the number of
mon species; each of these species was observed on 30 P. longicornis foragers across the four observation
different baits during the trial. Several species includ- times (F ⫽ 6.01; df ⫽ 3, 180; P ⬍ 0.005) and across the
ing the native Paratrechina vaga (Forel), Tapinoma sp., seven bait types (F ⫽ 17.1; df ⫽ 6, 60; P ⬍ 0.0001), and
and a third native, Pheidole sp., were not recorded the two effects were not independent (Ftime ⫻ bait ⫽
during the day trial in the same location. Pheidole 2.45; df ⫽ 18, 180; P ⬍ 0.0001). The posthoc analyses
megacephala (F.) was noticeably absent during the found consistent trends in preference with tuna,
night trial apart from its presence on one Maxforce Xstinguish, sugar water, and sugar water ⫹ boric acid
bait. preferred by P. longicornis throughout the 2-h trial
P. longicornis foragers also were recorded in low (Fig. 4).
numbers; not enough foragers were present to Þnd
signiÞcant bait preferences. During this trial they oc-
Discussion
curred at only 17 of the bait stations and only on 30
different baits. During the Þrst sampling round, there Bait Attractiveness. Of the seven baits available,
was an average of 6.6 P. longicornis foragers on 16 baits; tuna and Xstinguish were the most preferred by P.
during the second sampling round, an average of 7.7 longicornis foragers, followed by sugar water and sugar
foragers on 19 baits; and during the Þnal sampling water ⫹ boric acid. Wetterer et al. (1999) also found
round, an average of 10 foragers on 20 baits. The few that tuna (in oil) was consistently highly attractive to
foragers that were present were almost absent from P. longicornis during surveys in Biosphere 2, and Lee
the water, Maxforce, and Amdro baits, which is con- (2002) recommends sugar-based, liquid or gel formu-
sistent with the earlier trials, but they were also low in lations for control of P. longicornis. The granular baits
numbers on tuna in this trial. When considering all (Maxforce, Amdro, and Presto) were not successful at
ants from all species, there was signiÞcant concor- all in attracting P. longicornis foragers, particularly in
514 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 2

the trials where P. longicornis was the dominant ant guish bait used in management operations also would
species (trials 1 and 2). Although small numbers (⬍1% need to be quantiÞed for reptiles and other animals,
of active P. longicornis foragers) were observed on the which were attracted to the bait.
granular baits in the trials where other ant species The inßuence of competitive exclusion on the man-
were present (trials 3Ð5), it is likely that recruitment agement of P. longicornis incursions through the use of
was not sufÞcient to ensure adequate bait uptake and toxic baits will depend on the set of ant species present
spread through the colony, and no granules were at the incursion site and their relative competitive
observed being removed by P. longicornis foragers. abilities. P. longicornis is an aggressive, competitive
Lee and Tan (2004) report that baiting with paste species that can recruit large numbers of workers
and granular commercial baits is seldom effective rapidly to food resources (Clark et al. 1982, Lee 2002,
against P. longicornis in Singapore and Malaysia. Lester and Tavite 2004). In Tokelau, Lester and Tavite
High moisture content of baits is likely to be a key (2004) found that P. longicornis competitively ex-
factor in the attractiveness of baits to P. longicornis, cluded other ant species, including Tetramorium bi-
as with Argentine ant, Linepithema humile (Mayr), carinatum (Nylander) and Pheidole oceanica Mayr,
which is highly attracted to the moist matrix of from foraging on tuna baits. However, P. longicornis
Xstinguish (Klotz et al. 1996, Harris et al. 2002; R. can be displaced when more highly aggressive species,
Harris, unpublished data). such as S. invicta, recruit to the resource (Banks and
Other species of Paratrechina recruit well to Xstin- Williams 1989).
guish bait. The two undescribed Australian species of Application of Bait Preferences Trials to Effective

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

Paratrechina present in New Zealand have been ob- Management of P. longicornis Incursions. Xstinguish,
served foraging on the nontoxic version of Xstinguish tuna and sugar water (and sugar water ⫹ boric acid)
(Harris et al. 2002). Bait attractiveness trials on Palmyra were all highly preferred by P. longicornis. Although
Atoll showed Paratrechina bourbonica (Forel) had a tuna is attractive and may indicate a preference for
preference for sugar water followed by Xstinguish protein, it is not practical for use as a control tool in
(Krushelnycky and Lester 2003). P. bourbonica also terms of toxin incorporation and bait delivery. It is also
largely ignored Maxforce and Amdro granules in the highly attractive to nontarget animals, including birds
same study (Krushelnycky and Lester 2003). and reptiles. Sugar water is generally a highly pre-
Temporal Factors Affecting Recruitment. Repeat-
ferred food item to ants, but it is also a broad-spectrum
ing the summer bait attractiveness trial (no native ant
attractant. Care is required in situations where poten-
species co-occurrence), 6 wk later in autumn, did not
tial nontarget species, such as honey bees, Apis mel-
seem to affect either the abundance of P. longicornis
lifera L., are present. Bait delivery in liquid form is
foragers at the baits, or the relative attractiveness of
more difÞcult than solid bait delivery, and it requires
speciÞc baits. The temperature did decrease from
containerization, but it can be used efÞciently during
⬇30⬚C during the summer trial to ⬇23⬚C during the
autumn trial. small-scale incursion events (one or two nests). Xstin-
There was a reduction in P. longicornis foraging guish combines both a protein and carbohydrate
activity during the night trial (trial 4). However, be- (sugar) attractant (Stanley 2004) and was consistently
cause P. longicornis has been observed foraging in more attractive to P. longicornis than the sugar water.
large numbers throughout 24-h periods during sum- The attractiveness of Xstinguish to P. longicornis work-
mer at both these sites (W.A.R., unpublished data), it ers in this study conÞrms observations made during
is most likely that the reduction in foragers during this incursions in New Zealand (T. Ashcroft, personal
night trial was due to reduced reproduction in the communication). Because Xstinguish is produced as a
colony during late autumn, rather than reduced for- commercial ant bait, and it can be applied without the
aging at night. This was conÞrmed by reduced foraging need of bait stations or containers, it could be readily
activity during trial 5 (the next day), a late autumn used in both small and large-scale incursions.
repetition of the trial 3 diurnal trial. Post hoc analyses indicate that the presence of a
Effects of Co-Occurrence with Other Ant Species toxin in the sugar water (sugar water ⫹ 1% boric
on Bait Attractiveness. In regions where the native ant acid) did not deter P. longicornis foragers in these
fauna is depauperate and cryptic, native ant species trials. However, although never statistically signif-
would not be expected to exclude P. longicornis. Al- icantly different within a trial and a time, the sugar
though P. longicornis recruitment was substantially water with boric acid was always slightly less pre-
lower during trials where there was high native ant ferred than sugar water alone. Nevertheless, there
abundance and diversity (trials 3Ð5), there was still was always large numbers of foragers present on
reasonable recruitment to preferred baits (⬇50 for- sugar baits containing the boric acid. Exterm-An-
agers at each bait after 160 min). Ant (8% boric acid ⫹ 5.6% sodium borate in a sweet
Xstinguish was attractive to ant species other than solution) has been used to successfully attract large
P. longicornis. In areas where there are numerous con- numbers of P. longicornis foragers in previous in-
spicuous and highly abundant native ant species, the cursions at New Zealand ports (V. Van Dyk, per-
use of a toxic version of Xstinguish could have non- sonal communication), even though research indi-
target impacts. However, commercial baits, such as cates that concentrations of borates ⬎1% are
Amdro and Maxforce, proved to be similarly attractive repellent to some ant species (Klotz and Williams
to native ant species. Nontarget impacts of toxic Xstin- 1996, Hooper-Bui and Rust 2000). It is possible that
April 2007 STANLEY AND ROBINSON: ATTRACTIVENESS OF BAITS TO P. longicornis 515

the highly attractive nature of the Exterm-An-Ant References Cited

matrix overrides the repellent nature of boric acid.
Banks, W. A., and D. F. Williams. 1989. Competitive dis-
The effectiveness of boric acid-based baits in killing placement of Paratrechina longicornis (Latrielle) (Hyme-
P. longicornis colonies has not been tested. Although noptera: Formicidae) from baits by Þre ants in Mato
observations during New Zealand incursions suggest Grosso, Brazil. J. Entomol. Sci. 24: 381Ð391.
that Exterm-An-Ant is effective at controlling P. lon- Clark, D. B., C. Guayasamı́n, O. Pazmiño, C. Donoso, and Y.
gicornis (V. Van Dyk, personal communication), baits Páez de Villacı́s. 1982. The tramp ant Wasmannia au-
with high concentrations of boric acid (⬎1%) have ropunctata: autoecology and effects on ant diversity and
been shown to kill ants too rapidly and prevent re- distribution on Santa Cruz Island, Galápagos. Biotropica
cruitment (Klotz and Williams 1996, Hooper-Bui and 14: 196 Ð207.
Rust 2000). Baits with low concentrations of boric acid Collins, H. L., and A.M.A. Callcott. 1998. Fipronil: an ultra-
low-dose bait toxicant for control of red imported Þre ants
(ⱕ1%) are extremely effective at killing laboratory (Hymenoptera: Formicidae). Fla. Entomol. 81: 407Ð 415.
colonies of Monomorium pharaonis (L.), Tapinoma Davis, P., and J. J. van Schagen. 1993. Effective control of
melanocephalum (F.), S. invicta, and L. humile (Klotz pest ants. West. Aust. J. Agric. 34: 92Ð95.
and Williams 1996, Klotz et al. 1997, Ulloa-Chacon and Harris, R. J. 2002. Potential impact of the Argentine ant
Jaramillo 2003). Given the attractiveness of sugar wa- (Linepithema humile) in New Zealand. Department of
ter ⫹ 1% boric acid baits to P. longicornis in this study, Conservation, Wellington, New Zealand. (http://www.
the effectiveness of this bait in eliminating P. longi- conservation.org.nz/Publications/004⬃Science-and-
cornis colonies requires testing. Research/Science-for-Conservation/PDF/SFC196.pdf).

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016

Because of lack of registration in Australia, the Harris, R. J., J. S. Rees, and R. J. Toft. 2002. Trials to eradicate
infestations of the Argentine ant, Linepithema humile,
Xstinguish used in these trials was the nontoxic form (Hymenoptera: Formicidae) in New Zealand, pp. 67Ð74.
of the bait. Although the nontoxic form of Xstinguish In S. C. Jones, J. Zhai, and W. H. Robinson [eds.], Pro-
was highly attractive to P. longicornis foragers relative ceedings, The Fourth International Conference on Urban
to the toxic baits Amdro and Maxforce, it is possible Pests, 7Ð10 July 2002, Charleston, SC. Pocahontas Press,
that the toxic version of Xstinguish may repel P. lon- Blacksburg, VA.
gicornis. However, in numerous small- and large-scale Hedges, S. A. 1996a. Identical cousins. Pest Control Tech-
trials, no difference in the recruitment of Argentine nol. 24: 40 Ð101.
ants has been observed between the toxic and non- Hedges, S. A. 1996b. The great bait debate. Pest Cont. Tech-
toxic versions of the Xstinguish baits (Harris 2002; nol 24: 32Ð101.
Holway, D. A., L. Lach, A. V. Suarez, N. D. Tsutsui, and T. J.
Harris et al. 2002; V. Van Dyk, personal communica-
Case. 2002. The causes and consequences of ant inva-
tion). Furthermore, P. longicornis workers have been sions. Annu. Rev. Ecol. Syst. 33: 181Ð233.
observed to recruit well to the toxic version of Xstin- Hooper-Bui, L. M., and M. K. Rust. 2000. Oral toxicity of
guish bait during incursion events in New Zealand; no abamectin, boric acid, Þpronil, and hydramethylnon to
evidence of repellency has been observed (T. Ash- laboratory colonies of Argentine ants (Hymenoptera:
croft, personal communication). Hedges (1996a) re- Formicidae). J. Econ. Entomol. 93: 858 Ð 864.
ported P. longicornis would not feed for long enough Klotz, J. H., and D. F. Williams. 1996. New approach to
on commercial baits to ensure effective control. How- boric baits. IPM Practitioner 18: 1Ð 4.
ever, there was no evidence during this study that Klotz, J. H., D. H. Oi, K. W. Vail, and D. F. Williams. 1996.
Laboratory evaluation of a boric acid liquid bait on col-
foraging on preferred baits noticeably declined over
onies of Tapinoma melanocephalum, Argentine ants and
the 2Ð3-h trials. The effectiveness of Xstinguish in Pharoah ants (Hymenoptera: Formicidae). J. Econ. En-
eliminating P. longicornis colonies remains to be tomol. 89: 673Ð 677.
tested. The relative effectiveness of Xstinguish com- Klotz, J. H., K. M. Vail, and D. F. Williams. 1997. Toxicity of
pared with boric acid, sweet liquid baits also remains a boric acid-sucrose water bait to Solenopsis invicta (Hy-
untested. Given the attractiveness of Xstinguish to P. menoptera: Formicidae). J. Econ. Entomol. 90: 488 Ð 491.
longicornis in this study and the effectiveness of Xstin- Krushelnycky, P. D., and P. Lester. 2003. Report on obser-
guish at controlling Argentine ants (Harris et al. 2002) vations pertaining to ants and parasitic Hymenoptera,
and other ant species (V. Van Dyk, personal commu- Palmyra Atoll, November 18 Ð24, 2003. U.S. Fish and Wild-
life Service, Washington, DC.
nication), we predict high levels of control would be
Lee, C. Y. 2000. Performance of hydramethylnon- and Þpro-
achieved should efÞcacy trials be undertaken. nil-based containerised baits against household ants in
residential premises. Trop. Biomed. 17: 45Ð 48.
Acknowledgments Lee, C. Y. 2002. Tropical household ants: pest status, species
diversity, foraging behavior and baiting studies, pp. 3Ð18.
Zane Nicholls, Peter Collier, Elissa Terry, Jo Nicholls, and In S. C. Jones, J. Zhai, and W. H. Robinson [eds.], Pro-
Laura Adler assisted with Þeldwork. We thank Simon OÕConnor ceedings, The Fourth International Conference on Urban
(Biosecurity New Zealand), Travis Ashcroft (Biosecurity New Pests, 7Ð10 July 2002, Charleston, SC. Pocahontas Press,
Zealand), and Viv Van Dyk (Bait Technology Ltd.) for infor- Blacksburg, VA.
mation relating to ant incursion responses in New Zealand. Greg Lee, C. Y., and E. K. Tan. 2004. Guide to urban pest ants
Arnold provided advice on the statistical analyses, and Phil of Singapore. Singapore Pest Management Association,
Cowan, Richard Harris, Quentin Paynter, Richard Toft, and Singapore.
Darren Ward provided comments on earlier versions of this Lester, P. J., and A. Tavite. 2004. Long-legged ants, Anoplo-
manuscript. KingÞsher Bay Resort and Village (Fraser Island) lepis gracilipes (Hymenoptera: Formicidae), have in-
assisted with accommodation and transportation. This research vaded Tokelau, changing composition and dynamics of
was funded by Biosecurity New Zealand. ant and invertebrate communities. Pac. Sci. 58: 391Ð 401.
516 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 100, no. 2

SAS Institute. 1992. SAS technical report, SAS/STAT soft- on colonies of ghost ants (Hymenoptera: Formicidae). J.
ware: changes and enhancements, release 6.07. SAS In- Econ. Entomol. 96: 856 Ð 862.
stitute, Cary, NC. Wetterer, J. K., S. E. Miller, D. E. Wheeler, C. A. Olson, D. A.
Stanley, M. C. 2004. Review of the efÞcacy of baits used for Polhemus, M. Pitts, I. W. Ashton, A. G. Himler, M. M.
ant control and eradication. Landcare Research Contract Yospin, K. R. Helms, et al. 1999. Ecological dominance
Report LC0405/044. Landcare Research, Lincoln, New by Paratrechina longicornis (Hymenoptera: Formicidae),
Zealand. (http://www.landcareresearch.co.nz/research/ an invasive tramp ant, in Biosphere 2. Fla. Entomol. 82:
biosecurity/stowaways/Ants/bait_efÞcacy.asp). 381Ð388.
Ulloa-Chacon, P., and G. I. Jaramillo. 2003. Effects of boric
acid, Þpronil, hydramethylnon, and dißubenzuron baits Received 20 June 2006; accepted 2 January 2007.

Downloaded from http://jee.oxfordjournals.org/ by guest on March 22, 2016