HOUSEHOLD AND STRUCTURAL INSECTS
Bed Bug (Heteroptera: Cimicidae) Attraction to Pitfall Traps Baited
With Carbon Dioxide, Heat, and Chemical Lure
CHANGLU WANG,1,2,3TIMOTHY GIBB,2GARY W. BENNETT,2AND SUSAN MCKNIGHT4
J. Econ. Entomol. 102(4): 1580Ð1585 (2009)
Carbon dioxide (CO2), heat, and chemical lure (1-octen-3-ol and L-lactic acid) were
0.9% (mean ? SEM) of the bed bugs after 6 h, respectively. Traps baited with CO2? heat, CO2?
chemical lure, or CO2? heat ? chemical lure captured ?86.7% of the bed bugs after 6 h, indicating
baited pitfall traps were highly effective in attracting and capturing bed bugs from a short distance.
57.3 ? 6.4% of the bed bugs overnight. The pitfall trap was further tested in four bed bug-infested
apartments to determine its efÞcacy in detecting light bed bug infestations. Visual inspections found
an average of 12.0 ? 5.4 bed bugs per apartment. The bed bugs that were found by visual inspections
were hand-removed during inspections. A pitfall trap baited with CO2and chemical lure was
subsequently placed in each apartment with an average of 15.0 ? 6.4 bed bugs collected per trap by
the next morning. We conclude that baited pitfall traps are potentially effective tools for evaluating
bed bug control programs and detecting early bed bug infestations.
Cimex lectularius, lure, trap, monitoring
The recent resurgence of the bed bug, Cimex lectu-
larius L. (Heteroptera: Cimicidae), in the United
States, Canada, Australia, and some European coun-
tries triggered strong interests among researchers and
the pest control industry to investigate effective bed
bug management tactics (Cooper 2006, Gangloff-
bug infestations often go unnoticed until becoming a
serious problem. Once established, they are difÞcult
and expensive to eradicate due to insecticide resis-
Romero et al. 2007).
Detecting bed bugs during the early stages of an
bed bugs and will minimize the long-term manage-
ment cost. Unfortunately, bed bugs are difÞcult to
locate during visual inspections due to their small size
and cryptic behavior. A full inspection of an occupied
room usually requires two experienced technicians
and hours of labor. Even so, physical inspections are
Some pest management professionals have experi-
mented with sticky traps as a detection tool for bed
bugs and found them to be ineffective when used
alone (Cooper 2006). Lang et al. (2007) described a
system using a sticky surface and pitfall trap for trap-
ping bed bugs. However, there is no data demonstrat-
ing the effectiveness of the concept. Detection dogs
were found effective for identifying light bed bug
infestations (PÞester et al. 2008) but can be very ex-
a variety of factors including trainerÕs experience, dog
breed, and environmental factors.
Carbon dioxide (CO2), heat, and 1-octen-3-ol (oc-
tenol) are widely used in commercial traps to attract
to heat, odor, and CO2(Rivnay 1932, Marx 1955,
Aboul-Nasr and Erakey 1967). It is also known that
both CO2, octenol, and combinations of several short-
chained fatty acidsÑproprionic, butyric, valeric, and
L-lactic acidÑhave stimulating effect on the blood-
sucking bug Triatoma infestans (Klug) (Barrozo and
Lazzari 2004a,b). Anderson et al. (2009) investigated
bed bug response to pitfall traps baited with CO2
(emitted from a pressurized gas tank), thermal lure,
CO2was found to signiÞcantly and consistently in-
crease trap catches compared with traps without CO2
both in laboratory and Þeld experiments. Heat or lure
(combination of proprionic acid, butyric acid, valeric,
1Department of Entomology, Rutgers University, New Brunswick,
Entomology, Purdue University, West Lafayette, IN 47907.
3Corresponding author, e-mail: firstname.lastname@example.org.
4Susan McKnight, Inc., Memphis, TN 38111.
0022-0493/09/1580Ð1585$04.00/0 ? 2009 Entomological Society of America
acid, octenol, L-lactic acid) alone did not consistently
increase trap catches.
Based on results from Anderson et al. (2009), Bio-
watch) bed bug monitor. Around the same time, Cimex
Science LLC (Portland, OR) developed CDC3000 bed
lure to attract bed bugs. Yet, there are no experimental
data demonstrating the effectiveness of these traps and
Each device is estimated to cost several hundred to a
So far, our knowledge about bed bug response to
chemical and nonchemical lure is still very limited. In
chemical lure to bed bugs is poorly understood. The
objectives of this study were to 1) investigate the
ical lure) to bed bugs and 2) evaluate the utility of an
economic bed bug monitor for detecting low levels of
bed bug infestations.
Materials and Methods
nymphs) were collected from infested apartments in
Indianapolis, IN. They were maintained in glass jars
with folded Þlter paper as harborages. The bed bugs
were not fed during the study period. They were kept
at 22Ð23?C, 24Ð48% RH, and a photoperiod of 12:12
Experiment 1. Bed Bug Attraction to Pitfall Traps
Baited with CO2, Heat, and Chemical Lure. A pitfall
trap made of two plastic dishes was designed to eval-
uate bed bug attractant (Fig. 1A). The small dish size
Plastics Packing Corp., Bronx, NY). A layer of ßu-
oropolymer resin (DuPont Polymers, Wilmington,
DE) was applied to inner walls of the large dish and
was roughened with sand paper to allow bed bugs to
by width by height]) were used to evaluate bed bug
attractant (Fig. 1B). The bottom of each arena was
covered with fabric. A layer of ßuoropolymer resin
was applied to inner walls to prevent bed bugs from
fabric were placed at center of the arenas as harbor-
Twenty-Þve bed bugs that were collected from in-
fested apartments 7Ð8 d before the experiment were
released into each arena. Two traps were placed in
each arena (Fig. 1B). The bed bugs were conÞned to
the center of each arena under an 11-cm-diameter
the dark cycle, one trap in each arena received one of
the following three types of attractants: 1) CO2, 2)
heat, 3) chemical lure (1-octen-3-ol ? L-lactic acid)
The petri dish conÞning the bed bugs was removed.
Three arenas were placed in a room and separated by
2Ð3 m to minimize the confounding effects from the
attractants. The experiment was repeated three times
over three consecutive days. Each day, a different
attractant type was placed in each arena following a
Latin square design.
The CO2source was dry ice in double cups (a
236-ml foam cup inside a 355-ml insulated paper cup)
with a CO2release rate of 169 ml/min. CO2was re-
leased from a small oriÞce located on the lid of the
large cup. The heat source was a mini hand warmer
(Grabber,Grand Rapids, MI) placed inside the small
dish. Its surface temperature from 1 to 5 h after acti-
vation was 43.3Ð48.8?C. Air temperature immediately
at 21 h. The room temperature was 22.4?C during the
study period. The air temperature beside traps with
dry ice cup might have been negatively affected. The
ml micro centrifuge tube. Both chemicals were pur-
lid had a 2-mm-diameter opening to allow for slow
acid were effective in attracting bed bugs in our pre-
recorded hourly for 6 h and once at 21 h. The per-
centage of trapped bed bugs was calculated by divid-
of healthy bugs inside and outside the trap at the end
of the experiment. A healthy bed bug was deÞned as
tion to baited pitfall traps. (A) Pitfall trap. (B) Experimental
August 2009WANG ET AL.: BED BUG ATTRACTION TO BAITED PITFALL TRAPS
Experiment 2. Bed Bug Attraction to Pitfall Traps
binations (treatments) were evaluated under similar
conditions as experiment 1: 1) heat ? CO2; 2) heat ?
chemical lure; 3) CO2? chemical lure; and 4) heat ?
CO2? chemical lure. The CO2release rate was 83
ml/min. This rate was equally effective as that used in
experiment 1 in attracting bed bugs from a short dis-
tance. The objective of this experiment was to inves-
tigate whether combinations of two or three attract-
ants are more effective in attracting bed bugs and
which combination is most effective. The four treat-
ments were evaluated in four arenas on the same day.
Each treatment was assigned to a different arena and
the experiment was repeated four times over four
consecutive days following a Latin square design.
Each arena contained 30 bed bugs, which were col-
lected from infested apartments 7Ð19 d before the
experiment. They were conditioned for 24 h before
treatments were applied.
Experiment 3. Effectiveness of Baited Pitfall Traps
for Detecting Bed Bugs in Environmental Chambers.
height]; Van Ness Plastics, Clifton, NJ) was evaluated
for attracting and trapping bed bugs from long dis-
tances (Fig. 2). The trap design was the same as that
used by Anderson et al. (2009). A thin layer of talcum
powder (Spectrum Chemical Manufacturer Corp.,
Gardena, CA) was applied to inner surfaces of the
inverted cat feeding dish. Two mini-hand warmers,
two coffee mugs (390-ml capacity) Þlled with dry ice,
and a 0.7-ml tube of L-lactic acid and octenol (50 ?l
each, absorbed in cotton) were placed on top of the
This lasted for 8Ð9 h. The CO2release rate was in-
creased by a factor of Þve and the heat source was
doubled, compared with those in experiment 2. Be-
cause of the presence of dry ice, the air temperature
above the upper edge of the pitfall trap was 6.9?C
lower than the chamber temperature at the onset of
the experiment and was never higher than the cham-
ber temperature during the experiment period.
The experiment was conducted in four walk-in
chambers (3.05 by 1.83 by 2.67 m [length by width by
the ßoor-wall conjunction to prevent bed bugs from
climbing up the chamber walls. A thin layer of talcum
powder was applied to the plastic Þlm for added pre-
vention of escaping. The chambers were kept at 25Ð
Six pairs of harborages were placed at equal distances
along perimeters of the chamber ßoor. Each pair of
harborages included 3.5- by 3.5-cm folded cardboard
and 8- by 8-cm folded white cloth. Thirty bed bugs
were evenly released along perimeters of each cham-
ber. These bugs were collected from infested apart-
ments 3 d before the experiment.
After 1 d of acclimation in the chambers, a pitfall
trap was placed at the center of each chamber 1 h
before the dark cycle. The trap was examined after
21 h. Trapped bed bugs were counted and released
with healthy bed bugs. Each trap was then recharged
times over three consecutive days.
for Detecting Low Levels of Bed Bug Infestations in
Occupied Apartments. Utility of the pitfall traps as
described in experiment 3 was evaluated in four bed
bug-infested one-bedroom apartments. Each trap was
baited with CO2and chemical lure. Heat was not
provided because adding heat to the CO2? chemical
lure combination did not signiÞcantly improve the
trap catches in experiment 2. Three of the apartments
were treated with hot steam and/or insecticide spray
(0.5% chlorfenapyr) 2Ð4 wk before this experiment.
spray) 1 yr before this study. The apartments were
visually inspected for live bed bugs. Bed bugs that
were found during inspection were immediately re-
moved. Afterward, a trap was placed in the late after-
noon (3:30Ð4:30 p.m.) beside the infested bed or sofa
in each apartment and was examined the next morn-
ing. Residents stayed in the apartment, but not nec-
essarily in the same room with the trap. The visual
inspections found ?23 bed bugs in each apartment
for Detecting Bed Bug in an Unoccupied Apartment.
A heavily infested apartment which had ?500 bed
bugs by visual inspection was identiÞed. Most of the
bed bugs were found on a sofa in the living room. The
bedroom was nearly empty. The senior author of this
paper treated the apartment with hot steam and 0.5%
chlorfenapyr. The resident discarded the sofa and all
a week after treatment. At 13 and 21 d after vacancy,
a baited trap was placed where the sofa was located.
The trap was examined the next morning.
Data Analysis. A paired t-test was used for compar-
ing trap catches between baited and nonbaited traps.
Analysis of variance (ANOVA) was used to compare
Baited pitfall trap for monitoring bed bug pop-
1582JOURNAL OF ECONOMIC ENTOMOLOGY
Vol. 102, no. 4
percentage of trap catches among different treat-
ments. This was followed by TukeyÕs honestly signif-
of bed bugs trapped in the pitfall traps over time
(days) was analyzed using repeated measurement
analysis. All analyses were performed using SAS soft-
ware (SAS Institute 2003).
Bed Bug Attraction to Pitfall Traps Baited with
CO2, Heat, and Chemical Lure. Bed bugs were de-
of the trap catches occurred within 6 h (Fig. 3A).
Traps baited with CO2, heat, and chemical lure cap-
bugs from the arenas at 6 h, respectively. In contrast,
the accompanying nonbaited traps trapped 3.0 ? 3.0,
16.0 ? 5.3, and 31.3 ? 11.7% of the bed bugs, respec-
nonbaited traps (P ? 0.05). Chemical lure did not
0.48). CO2was signiÞcantly more attractive to bed
bugs than heat (F ? 17.2; df ? 2, 6; P ? 0.003).
In experiment 2 evaluating bed bug attraction to
four combinations of attractants, the mean percent-
ages of bed bugs in the nonbaited trap were ?2.5%.
heat ? CO2, heat ? chemical lure, CO2? chemical
72.1 ? 6.6, 88.7 ? 3.5, and 89.0 ? 1.0%, respectively
(Fig. 3B). All combinations containing CO2were sig-
niÞcantly more attractive to bed bugs than that with-
out CO2(P ? 0.05; TukeyÕs test) and attracted bed
bugs much faster than CO2alone.
Effectiveness of Baited Pitfall Traps for Detecting
Bed Bugs in Environmental Chambers and in Apart-
ages of bed bugs (average of 3 d and four chambers)
captured by baited traps were 15.1 ? 2.5, 41.8 ? 7.4,
and 57.3 ? 6.4% at 2, 6, and 22 h after placement,
respectively (Fig. 4). There were no signiÞcant dif-
2, 6, or 22 h (P ? 0.05; ANOVA).
In occupied apartments, baited pitfall traps cap-
tured bed bugs that were not detected by visual in-
of bed bugs detected by trapping and by visual in-
spection were 15.0 ? 6.4 and 12.0 ? 5.4, respectively.
In the unoccupied apartment, the baited pitfall trap
captured 505 bed bugs (91.1% were nymphs) on 13 d
after vacancy and 113 bed bugs on 21 d after vacancy.
We have demonstrated that baited pitfall traps can
can be used to help evaluate effectiveness of bed bug
control programs. Pitfall traps baited with CO2and
chemical lure captured bed bugs that were not dis-
covered by visual inspections in apartments. A visual
to pitfall traps baited with a single attractant (A) and a
combination of two or three attractants (B).
Percentage (mean ? SEM) of bed bugs attracted
Thirty bed bugs were in each walk-in chamber every day.
Daily changes in trap catches (mean ? SEM).
and CO2for detecting bed bugs in apartments
Effectiveness of pitfall traps baited with chemical lure
One trap was placed in each apartment after visual inspection and
hand-removal of bed bugs found by visual inspection.
August 2009WANG ET AL.: BED BUG ATTRACTION TO BAITED PITFALL TRAPS
preparing and examining baited pitfall traps require
little effort and are much more efÞcient than visual
inspection, especially in multiunit buildings where
many apartments need to be inspected in a short
period of time.
an efÞcient method to quantify the attractiveness of
nonchemical and chemical lures to bed bugs. Among
the three attractants (CO2, heat, and chemical lure)
in attracting bed bugs, supporting the conclusion by
Anderson et al. (2009). Additionally, we found that
ical lure increased trap catches in two of the three
replicates at 6 h and in all replicates at 21 h. However,
the differences were small. This result was consistent
with our preliminary tests showing weak attractive-
octenol might have negatively affected each otherÕs
release rate. The synergistic effect between CO2and
other lures was not analyzed statistically because the
adding heat or chemical lure signiÞcantly improved
the rate and speed of trap capture.
In environmental chamber studies, we initially
tested the small pitfall traps for 1 d. Only 18.1 ? 7.3%
bed bugs were captured. We then used a much larger
trap design, higher CO2release rate, and doubled the
for the much larger experimental area compared with
experiment 2. Even so, a much lower percentage of
bed bugs was trapped compared with experiment 2
(57.3 versus 89.0), suggesting that both distance be-
tween bed bugs and the trap and lure release rate or
exactly the same inverted cat feeder, Anderson et al.
were captured overnight by a pitfall trap baited with
CO2(released from a pressurized tank) in a 1.83-m2
capturing bed bugs. This is important because dry ice
is much more available and affordable than pressur-
ized CO2tanks or cartridges. The hand warmers were
not able to generate enough heat to raise the air
temperature above the trap when dry ice cups were
trap catches was dubious in this design and needs to
the mean number on day 3 was much larger than that
differences. During the 3-d experiment, an average
27.2 ? 2.9% bed bugs died each day, indicating that
many of the tested insects might have suffered from
injury. Yet, the trap performance was consistent over
the 3-d experiment period, validating the reliability of
Anderson et al. (2009) documented the perfor-
mance of baited pitfall traps both in occupied and
ulation levels were not determined. Thus, the relative
efÞcacy of the trap was not clear. Our study in apart-
ments with known bed bug population levels demon-
strated that pitfall traps were able to detect low level
bed bug infestations, even in the presence of a human
or in different rooms). All traps caught bed bugs that
were not found by visual inspections, showing that
small numbers of bed bugs were still present, and
competition from human host did not inhibit pitfall
traps from detecting low numbers of bed bugs.
Bed bugs can live without a human host for an
extended period. In unoccupied apartments where
bed bugs are still present, the effectiveness of baited
pitfall traps may be higher because bed bugs are hun-
test in a vacant apartment showed that in heavily
infested apartments 1) visual inspections could seri-
ously underestimate the bed bug numbers, 2) large
numbers of bed bugs were not on the furniture and
survived the chemical and nonchemical treatment,
and 3) baited-pitfall traps were helpful in monitoring
effectiveness of bed bug treatments. From our obser-
vations, bed bugs frequently travel from infested
apartments to the hallways in a multiunit apartment
infer that bed bugs are more likely to disperse into
neighboring apartments through hallways when their
in unoccupied infested apartments may reduce the
risk of bed bug dispersal between adjacent units in
ecology is required for optimizing the trap design and
using traps in a way that maximizes efÞcacy. Consid-
control (Cooper 2006, Romero et al. 2007, Potter
2008), and the continuing spread of bed bug infesta-
tions in the United States (Gangloff-Kaufmann et al.
2006), an effective and reliable bed bug monitoring
tool will probably play a pivotal role in safeguarding
human health, reducing insecticide applications, and
minimizing economic losses associated with bed bug
We thank Mahmoud Aboul El-Nour and Eva Chin for
technical assistance; Indianapolis Housing Agency staff for
providing access to apartments; Lei Shu for statistical assis-
tance; and Jill Gordon, Richard Cooper, and two anonymous
reviewers for reviewing an earlier draft of the manuscript.
This is New Jersey Experiment Station publication D-08-
Aboul-Nasr,A.E.,andM.A.S.Erakey. 1967. Onthebehavior
and sensory physiology of the bed-bug. I. Temperature
reactions (Hemiptera: Cimicidae). Bull. Soc. Entomol.
Egypte 51: 43Ð54.
1584JOURNAL OF ECONOMIC ENTOMOLOGY
Vol. 102, no. 4