A lethal ovitrap-based mass trapping scheme for dengue control in Australia: I. Public acceptability and performance of lethal ovitraps


Scott Ritchie, School of Public Health, Tropical Medicine and Rehabilitation Sciences, James Cook University, PO Box 6811, Cairns, QLD 4870, Australia. Tel.: + 61 740503619; fax: + 61 740311440; e-mail: scott.ritchie@jcu.edu.au


We report on the first field evaluation of the public acceptability and performance of two types of lethal ovitrap (LO) in three separate trials in Cairns, Australia. Health workers were able to set standard lethal ovitraps (SLOs) in 75 and 71% of premise yards in the wet and dry season, respectively, and biodegradable lethal ovitraps (BLOs) in 93% of yards. Public acceptance, measured as retention of traps by residents, was high for both trap types, with <9% of traps missing after 4 weeks. Traps retaining water after 4 weeks were 78 and 34% for the two SLO trials and 58% for the BLOs. The ‘failure rate’ in the 535 BLOs set in the field for 4 weeks was 47%, of which 19% were lost, 51% had holes from probable insect chewing, 23% were knocked over, 7% had dried by evaporation and 1% were split. There was no significant difference in the failure rate of BLOs set on porous (grass, soil and mulch) versus solid (tiles, concrete, wood and stone) substrates. The SLOs and the BLOs were readily acceptable to ovipositing Aedes aegypti L. (Diptera: Culicidae); the mean number of eggs/trap was 6 and 15, for the dry season and wet season SLO trial, respectively, and 15 for the BLO wet season trial. Indeed, 84–94% of premise yards had egg positive SLOs or BLOs. A high percentage of both wet and dry season SLOs (29 and 70%, respectively) and BLOs (62%) that were dry after 4 weeks were egg positive, indicating the traps had functioned. Lethal strips from SLOs and BLOs that had been exposed for 4 weeks killed 83 and 74%, respectively, of gravid Ae. aegypti in laboratory assays. These results indicate that mass trapping schemes using SLOs and BLOs are not rejected by the public and effectively target gravid Ae. aegypti. The impact of the interventions on mosquito populations is described in a companion paper.


Dengue fever is endemic throughout tropical and subtropical regions of the world, where up to 2.5 billion people live at risk from epidemic transmission and an estimated 50–100 million cases occur annually (Gubler, 1998; Farrar et al., 2007). In north Queensland, Australia, dengue is not endemic, but outbreaks, generally coinciding with the wet season (January to April), are an almost annual occurrence (Ritchie et al., 2002; Hanna et al., 2006). From January 2000 to November 2008 there have been 16 outbreaks of dengue, resulting in 1142 confirmed cases (Queensland Health, unpublished data), two of which were fatal (McBride, 2005).

Aedes aegypti L., the primary vector of dengue viruses, feeds almost exclusively on humans, rests in sheltered locations within the home and relies on water held in artificial containers for the development of immature stages (Christophers, 1960). Aedes aegypti is found in many urban centres in north Queensland, and container types utilized vary with local conditions, but in Cairns, northern Queensland, the most productive containers include discarded household items, tarpaulins, pot plant bases, sump pits, roof gutters, rubbish and garden accoutrements (Montgomery & Ritchie, 2002; Montgomery et al., 2004; Hanna et al., 2006). Suburbs featuring open, elevated and unscreened ‘Queenslander’-style houses have been subject to larger outbreaks (e.g. Hanna et al., 2006).

In Cairns, control of dengue outbreaks is achieved under the Dengue Fever Management Plan for North Queensland 2005–2010 (http://www.health.qld.gov.au/dengue/) using, in part, coordinated and sustained vector control conducted by a specialized Dengue Action Response Team (DART) of Queensland Health (Ritchie et al., 2002). The DART has traditionally relied on a combination of selective interior residual spraying (SIRS) and larval control to reduce Ae. aegypti populations. While indoor application of residual insecticides has been shown to enhance dengue control (Reiter & Gubler, 1997; Hanna et al., 2001; Perich et al., 2001; Ritchie et al., 2002), it is not always publicly acceptable because of concern with pesticides in domestic environments (Ritchie, 2005). Interior spraying is laborious, making it difficult to cover large areas affected by explosive dengue outbreaks (Ritchie, 2005; Hanna et al., 2006). Furthermore, the elimination of oviposition sites can induce infective mosquitoes to fly beyond the treatment area in search of oviposition sites, and hence extend the distribution of dengue transmission (Reiter et al., 1995; Edman et al., 1998). Clearly a dengue control strategy that is rapid while minimizing pesticide use is needed.

To combat the concerns associated with indoor pesticide applications, Queensland Health has initiated the strategic use of traps that ‘Lure and Kill’ (L&K) the adult mosquito while minimizing exposure of non-target organisms to insecticides. The L&K strategy relies on the use of a chemical lure (kairomone) that selectively attracts the target organism and a lethal target or trap that subsequently kills or collects the attracted animals. Use of L&K traps/targets in mass trapping schemes has been successful in the control of many agricultural pests (Ridgway et al., 1990; Nalyanya et al., 2000), tsetse flies (Vale, 1993) and, to a limited extent, mosquitoes (Kline, 2006).

The lethal ovitrap (LO) is the cornerstone of L&K trapping for dengue control. Zeichner & Perich (1999) conducted the first laboratory research into making the regular ovitrap (RO), used by many countries as a surveillance tool, lethal to Ae. aegypti. The wooden paddle, provided as a substrate for mosquito oviposition, was replaced with a velour paper strip treated with a synthetic pyrethroid that would kill mosquitoes attempting to oviposit. In small cage trials using deltamethrin, they achieved 98% control of female Ae. aegypti. This deltamethrin LO has been successfully used in field trials against Ae. aegypti in Thailand (Sithiprasasna et al., 2003) and Brazil (Perich et al., 2003).

In Cairns, vector control officers use LOs that have an oviposition strip made from a red flannel cotton strip that has been treated with the synthetic pyrethroid bifenthrin, here referred to as a standard lethal ovitrap (SLO) (Ritchie, 2005; Williams et al., 2007). Williams et al. (2007) demonstrated that these SLOs were highly successful, killing 92% of ovipositing mosquitoes in small cage trials, and that they were still effective after a 4-week deployment in the field with no significant loss of bifenthrin or lethal toxicity to Aedes. However, LOs constructed from plastic containers (e.g. Zeichner & Perich, 1999; Williams et al., 2007) have the potential to produce mosquitoes after the insecticide degrades. Thus, SLOs must be labelled and archived into a database for timely retrieval, and having to set and retrieve SLOs creates a doubling of the workload. To counteract this, Ritchie et al. (2008) developed a biodegradable lethal ovitrap (BLO), similar in appearance and dimensions to the plastic SLO (Fig. 1), and composed of plasticized amylose maize polymers that slowly degrade upon contact with water. Ideally, BLOs fail structurally (i.e. breakdown enough to leak), losing water before the insecticide degrades and mosquito breeding occurs, thus they do not need retrieval, providing considerable savings in logistics and labour.

Figure 1.

Standard lethal ovitrap (left) and biodegradable lethal ovitrap used in the study.

No studies of public acceptability and performance of SLOs or BLOs in large-scale controlled field situations have been published, and the acceptance and fate of the SLOs and BLOs used in Cairns have not been assessed in a large controlled field situation. In the present study, we present the results of a mass-deployment of SLOs and BLOs that indicates the acceptability of both traps to the public. We also measured trap performance in terms of trap structural failure, mosquito oviposition and trap lethality. In a companion paper, we quantify the affect of SLO and BLO interventions on Ae. aegypti populations in north Queensland (Rapley et al, 2009).

Materials and methods

Standard lethal ovitrap intervention

Two interventions using SLOs were conducted during the dry (June to July 2006) and wet (March to April 2007) seasons in Machans Beach, a suburb of Cairns, Queensland, Australia. This suburb generally has large populations of Ae. aegypti and in 2000 had 49 confirmed cases of dengue (Ritchie et al., 2001). Daily rainfall, temperature and evaporation were obtained from an Australian Bureau of Meteorology recording station located ∼1 km from the study site.

This paper describes only the acceptability and performance of SLOs and BLOs, whereas the treatments and population assessments will be described fully in the companion paper (Rapley et al, 2009). Public acceptability was not directly measured by interview or questionnaire. Rather, instances where traps were not disposed of by residents were interpreted as public acceptance. In each season, a geographic information system (GIS; MapInfo, PitneyBowes, North Sydney, Australia) was used to define two comparably-sized (∼12 ha) oval treatment enclaves, one for ‘larval control’ and one for ‘larval control’ plus SLOs. Larval control involved reducing actual and potential Ae. aegypti immature habitats by turning over both wet and dry containers, and treating permanent wet containers and roof gutters with s-methoprene pellets. This was conducted only once, concurrently with the setting of the SLOs or BLOs. SLOs (Fig. 1) consisting of a 1.2-L black plastic bucket with a 13.5 by 5.0 cm bifenthrin-treated red flannel ovistrip (7 mg AI/strip) were filled with ∼1 L of tap water (Williams et al., 2007). A 0.5-g alfalfa pellet was added to create an infusion attractive to gravid Ae. aegypti (Ritchie, 2001). To reduce degradation of the pesticide strip, SLOs were placed in secluded and sheltered sites protected from rainfall outside houses and buildings according to Williams et al. (2006). The dry and wet season treatment enclaves receiving SLOs contained 77 and 96 houses, respectively. Larval control and setting of SLOs was only done with the permission of residents. No residual sprays or adulticidal fogging that could have confounded results were conducted.

Standard lethal ovitrap acceptability and performance

After 4 weeks deployment, the SLOs were collected and scored for the presence or absence of water. Missing or destroyed traps were recorded. Lethal ovistrips were removed and mosquito eggs counted under a stereo-microscope. Ovistrips were allowed to dry for 2 days before their viable eggs were picked from their surface using a scalpel blade, into a hay infusion to induce hatching (Williams et al., 2007). Larvae were reared to fourth instar and identified. The number of Ae. aegypti eggs was calculated by adjusting the total number of eggs oviposited on the ovistrips for the presence of eggs of other container breeding mosquitoes.

Ovistrip lethality

To ensure the ovistrips were still lethal after 4 weeks of field exposure, we exposed groups of 10 female Ae. aegypti (Cairns colony; F3–6) to strips using a modified WHO cone (Williams et al., 2007). We compared the mean mortality of mosquitoes exposed to field-deployed lethal ovistrips (n = 15), fresh lethal ovistrips (n = 6) and an untreated flannel strip (control; n = 6), using the Mann–Whitney U-test (Zar, 1999). This was only done for SLOs from the wet season intervention.

Biodegradable lethal ovitrap intervention

The Cairns suburb of Parramatta Park was chosen as the experimental site because of its high Ae. aegypti abundance and past occurrence of dengue (Ritchie et al., 2004; Hanna et al., 2006). Temperature and evaporation data were obtained from an Australian Bureau of Meteorology recording station located ∼5 km from the study site, whereas rainfall was measured <1 km from the site. Using GIS (MapInfo), three comparably sized geographical areas were designated as treatment areas. Three teams of three people implemented the treatments over 3 days (wet season, 18–20 February, 2008). The intervention involved the mass deployment of BLOs (usually 4/house as per Williams et al., 2006) and larval control (as described above) when traps were set. Larval samples were collected from containers, identified and the Breteau Index (BI) calculated; this was done when traps were set (pre-treatment) and 4 weeks later (post-treatment). The BLOs (Ritchie et al., 2008) consisted of a 1.2-L biodegradable thermoplastic starch bucket, with a 13.5 by 5 cm bifenthrin-treated ovistrip and filled with 1 L of tap water. Plastic netting normally placed over SLOs to prevent litter from falling into the traps (Fig. 1), was not used on the BLOs as it is not biodegradable. The BLOs were set outside houses in secluded and sheltered sites similar to as described by Williams et al. (2006) for SLOs and sticky ovitraps (SOs), and the type of substrate beneath the BLOs was recorded. The BLOs were deployed for a period of 4 weeks (19 February to 20 March 2008) and, in contrast to the SLO intervention, the new Queensland Health Act (2005) enacted in 2007 enabled us to set BLOs and conduct larval control using methoprene in private yards when residents were not at home.

Biodegradable lethal ovitrap acceptability and performance

At the end of the 4 weeks, the BLOs were collected and information relating to their field performance was recorded. This information included whether the trap was still holding water, had any type of damage or had disappeared. Traps were returned to the laboratory where ovistrips were removed, eggs counted under a stereo-microscope and the ovistrips were scored for per cent coverage (<25, 25–50, 50–75, >75%) with fungi or gelatinous film that might deter mosquito contact. The percentage of eggs that were Ae. aegypti was estimated using the percentage of Aedes collected on sticky ovitraps (Ritchie et al., 2003; n = 20/intervention area/week for a total of 240) deployed in the area during the intervention. The inner surface of a random sample (n = 89) of BLO buckets was also examined for eggs using a stereo-microscope to determine if the lethal ovistrip was being avoided by ovipositing females.

Ovistrip lethality

The lethality of field-exposed bifenthrin-treated ovistrips was compared with untreated ovistrips using a small cage choice trial (Zeichner & Perich, 1999). This procedure was chosen because it provided the mosquitoes with a choice of oviposition sites, and also allowed a measure of oviposition on the ovistrips. The untreated control ovistrip was not aged in the field, and this may have impacted results. Ten gravid female Ae. aegypti were placed into a 0.10-m3 cage containing a regular ovitrap (RO) (a 1.2-L black plastic bucket with a untreated flannel cloth ovistrip containing 1 L of tap water and a 0.5-g lucerne pellet, according to Ritchie, 2001), and a randomly selected BLO that had been deployed for the 4-week field trial (n = 18). A cotton pad soaked in 10% sugar water was provided for food. A similar cage with two ROs containing untreated ovistrips served as a control (n = 10). The number of dead females was measured at 48 h and compared between treatment and control with the Mann–Whitney U-test (Zar, 1999). The mean number of eggs/ovistrip was compared using an unpaired t-test.


Acceptability and fate of field-deployed lethal ovitraps

Standard Lethal Ovitraps. In total, 206 and 243 SLOs were set at 54 and 64 premises during the dry (3–31 July 2006) and wet (26 March to 23 April 2007) season SLO interventions, respectively. Public acceptance of the SLOs was high in both interventions (Table 1). In the dry and wet season SLO interventions, we were denied entry to only five and six houses, respectively. In a further 13 (25%) and 20 (29%) houses, respectively, we were denied permission by residents to set SLOs (residents' decisions were accepted without further questioning), but they granted us permission to perform larval control. During both SLO interventions, only a small number of SLOs were lost (i.e. disappeared) in situ, indicating a low rejection rate of set traps by the public. A noteworthy percentage of traps dried out before the end of the intervention period, especially during the wet season trial (Table 1), reflecting higher temperatures in the wet season and unusually dry conditions at the time. Only 77 mm of rain fell from 26 March to 23 April 2007. The warm conditions (mean daily minimum and maximum temperature of 21.2 and 29.5°C, respectively), allowed for high total evaporation of 169 mm and drying of the SLOs. While it was drier during the dry season intervention (rainfall = 41 mm), cooler weather (mean daily minimum and maximum temperature of 17.7 and 25.6°C, respectively) restricted evaporation to 125 mm.

Table 1.  Public acceptance and fate of standard (SLO) and biodegradable (BLO) lethal ovitraps after a 4-week field deployment.
ParameterStandard lethal ovitrap (SLO)Biodegradable lethal ovitrap (BLO)
Dry seasonWet seasonWet season
  1. *This trial took place after Queensland Health Act enacted that allowed for SLOs and BLOs to set in yards without permission.

Number of houses in intervention area 72 90179
Number (%) houses denied access 5 (7) 6 (7) 13 (7)
Number (%) of houses where trapping was possible 54 (75) 64 (71)166 (93)*
Number of traps set206243553
Number (%) of traps recovered204 (99)235 (97)505 (91)
Number (%) of traps lost/missing 2 (1) 8 (3) 48 (9)
Number (%, recovered traps only) of wet traps159 (78) 80 (34)294 (58)
Number (%, recovered traps only) of dry traps 45 (22)155 (66)211 (42)

Biodegradable lethal ovitraps

We were able to set BLOs and conduct larval control in 93% of premises (Table 1). In the remaining 7% of premises (13) entry was denied by the resident or we were unable to access the yard as a result of locked gates or aggressive dogs. Data for the three treatment areas were pooled. Overall, a total of 553 BLOs were set in 166 premises, with a mean of 3.3 per premise. After being deployed in the field for 4 weeks (19 February to 20 March 2008), 53% of the BLOs were holding water (Table 2). Weather during the intervention was hot and wet; mean daily minimum and maximum temperatures were 23.4 and 29.7°C, respectively, and precipitation totalled 1114 mm. Total evaporation was 158 mm. Only 9% of the traps were lost, indicating few traps were removed despite many being set without consent. Excluding the 48 lost traps, 211/505 (42%) recovered BLOs were not holding water (Table 2). The overall failure rate (% of total set) was 47%, as a result of animal chewing (24%), knocked over (11%), drying out (3%) and splitting (1.5%). Adjusting for the number of recovered traps, this percentage increased slightly (Table 2). Indeed, only 26% (133) of the 505 recovered BLOs structurally failed, with the remainder having been knocked over or dried out. There was no significant difference in the failure rate of BLOs set on solid compared with porous substrates (χ2 = 0.16, P = 0.69, d.f. = 1; Table 3). We suspect that cockroaches chewed most of the holes found in the BLOs (Fig. 2), although we have also found ants, isopods and amphipods feeding on the BLOs. While chewing was a problem (311 or 62% of recovered BLOs had chew marks), only 131 of these (42%) had failed within the 4-week intervention. Most of the BLO lethal ovistrips were free of fungus or gelatinous films after 4 weeks. A total of 499 ovistrips scored for per cent contamination by fungus, debris or gelatinous film was: <25, 80.0; 25–50, 12.8; 50–75, 2.4 and >75, 4.8%.

Table 2.  Causes of failure in biodegradable lethal ovitraps (BLOs) deployed for 4 weeks in the field.
  1. *Excludes missing traps.

Total traps deployed553 
Total number of traps failed (% of total deployed)259 (47) 
Number of failed traps recovered* (% of total failed)211 (81) 
Reason for failureNumber% of total traps failed% of total failed traps recovered*
Trap missing 4819not applicable
Trap chewed traps1315162
Trap knocked over 592328
Trap evaporated dry 19 7 9
Trap split 2 1 1
Table 3.  Failure rate of biodegradable lethal ovitraps (BLOs) set on solid and porous substrates for 4 weeks.
SubstrateBLOs setBLOs failedFail (%)
  1. *Includes leaves (4), metal (3), carpet (2) and plastic (2).

Sum solid substrates33313239.6
Sum porous substrates1617546.6
All substrates50521342.2
Figure 2.

Failed biodegradable lethal ovitrap bucket with chew marks and holes, probably caused by cockroaches.

Oviposition in standard lethal ovitraps and biodegradable lethal ovitraps

Standard lethal ovitraps

A greater percentage of SLOs were positive for mosquito eggs during the wet season than the dry season (Table 4). Over half of the recovered SLOs and BLOs had eggs on the lethal strips for both interventions. Most houses had at least one positive SLO, with 86 and 97% of premises having at least one positive SLO after the dry and wet season interventions, respectively. While wet traps had more eggs than dry traps, and significantly so during the wet season trial, many dry SLOs had eggs and had probably killed mosquitoes. All larvae hatched from a sample of eggs (dry season n = 98, wet season n = 157) collected on bifenthrin-treated ovistrips were Ae. aegypti, therefore the number of adults killed could be determined (Williams et al., 2007) without having to adjust egg counts for the presence of other species.

Table 4.  Oviposition in standard lethal ovitraps (SLOs) and biodegradable lethal ovitraps (BLOs) after a 4-week field deployment.
ParameterStandard lethal ovitrap (SLO)Biodegradable lethal ovitrap (BLO)
Dry seasonWet seasonWet season
  1. *Denotes that the mean number of eggs/lethal ovistrip in wet and dry traps is significantly different using unpaired t-test.

Number of traps recovered204235505
Number (%) with eggs112 (55)171 (73)294 (58)
Mean ± SD number of eggs (all traps)5.8 ± 10.112.0 ± 16.313.2 ± 22.7
Number of wet traps15980294
Number of dry traps45155211
Number of wet traps with eggs (%)93 (59)62 (78)183 (62.2)
Number of dry traps with eggs (%)13 (29)109 (70)110 (52.1)
Mean ± SD * number of eggs (wet traps)6.2 ± 10.515.4 ± 19.4*14.6 ± 28.4
Mean ± SD number of eggs (dry traps)4.3 ± 8.410.3 ± 14.1*11.3 ± 24.0

Biodegradable lethal ovitraps

The BLOs were readily accepted by mosquitoes, with 58% positive for Aedes eggs (Table 4). As with the SLOs, most houses (84%) had at least one positive BLO. Premises with Ae. aegypti breeding at the end of the intervention (n = 31) were no more likely to have positive BLOs (90.3%) than those without breeding (87.9%; χ2 = 0.0007, P = 0.98, d.f. = 1), indicating that competition from normal receptacle breeding sites was insignificant. Of the 259 dry traps, 52% were egg positive (Table 4), and the mean number of eggs/lethal ovistrip was comparable (t = 0.899; P = 0.269) to that found in wet BLOs, suggesting structural failure was relatively recent. Off-strip oviposition was checked in 89 BLOs, and only 7/89 (8%) were positive (for a total of 13 eggs), indicating strong oviposition preference for the flannel lethal ovistrip. Almost all Aedes collected on the sticky ovitraps were Ae. aegypti (171/173, 99%), with the remainder being Aedes palmarum Edwards. Thus, we assumed that all eggs on the BLO lethal ovistrips were Ae. aegypti.

Toxicity of the lethal ovistrips

The bifenthrin-treated ovistrips deployed in SLOs for 4 weeks in the wet season intervention at Machans Beach still killed in excess of 80% of mosquitoes exposed to the lethal ovistrip for 4 min (Fig. 3). However, mean percentage kill was significantly lower compared with unexposed bifenthrin-treated ovistrips (Z = 2.55; P = 0.01). Oviposition choice bioassays were conducted on 14 and 4 lethal ovistrips with <25% and 25–50% fungal contamination, respectively, collected from BLOs set in the field for 4 weeks. These lethal ovistrips killed a significantly higher mean (±SE ) proportion of gravid Ae. aegypti (0.73 ± 0.04) than the untreated control strips (0.04 ± 0.01) (Z = 4.21, P < 0.001). A mean (±SE ) of 259 ± 26 and 117 ± 12 eggs were laid on the control and lethal ovistrips, respectively, a highly significant difference (t = 4.24, P < 0.001).

Figure 3.

Mean (±SE) percentage mortality of groups of 10 female Aedes aegypti exposed for 4 min to bifenthrin-treated ovistrips that were not used in the field (n = 6, unexposed control), bifenthrin-treated ovistrips that were set in standard lethal ovitraps at Machans Beach for 4 weeks (n = 15, field-exposed) and untreated ovistrips (n = 6, untreated control).


This is the first published account documenting the public acceptance of lethal ovitraps, and we found a high level of public acceptance for use of both standard and biodegradable models. We realize that acceptability is used in a broad sense. Residents were not interviewed and their opinions were not sought. Rather, the failure of residents to remove SLOs and BLOs set in their yards is used as a proxy for acceptance. While this does reflect ‘realistic use’, whereby traps would be set in yards either with permission or, if no one was home, without formal permission, further studies should be conducted to accurately gauge public opinion of the trapping scheme. Interestingly, the introduction of the new Queensland Public Health Act (2005) enabled us to set BLOs in a greater percentage of houses (93%) than occurred when permission was required during the two SLO interventions (75 and 71%). We were refused entry into very few properties; reasons for not being able to set SLOs included permission being flatly denied (various motives), properties being vacant, aggressive dogs in yards and locked gates. Of the SLOs and BLOS set, very few (<5%) were removed by residents. Thus, an SLO and BLO mass trapping scheme is a feasible dengue intervention with regard to the public not interfering with or rejecting the traps.

For both SLOs and BLOs, about half of the traps were still wet after the 4-week intervention period. Many of the dry traps were egg positive, indicating that they had successfully attracted and likely killed gravid Ae. aegypti. However, drying obviously reduced the effectiveness of SLOs towards the end of the intervention period. This was particularly a problem in the wet season, as a result of the higher total evaporation during this period. The high number of traps found without water at the end of the wet season intervention indicates that the trap in its current form may be more effective if it is replaced or re-filled with water after 3 weeks of operation, as opposed to the 4-week test period and recommended by Williams et al. (2007). In drier areas, such as Townsville to the south of Cairns, SLOs can dry out within 2 weeks, and have to be refilled (G. Ehler, Queensland Health, personal communication), although the use of larger buckets improves SLO water retention.

Both SLOs and BLOs were acceptable to ovipositing Ae. aegypti. Most houses had positive traps, indicating that at least some level of control was being achieved throughout the intervention area. In the case of the BLO intervention, the presence of containers with immature Ae. aegypti that might compete with the BLOs did not reduce the percentage of BLOs that were positive. Skip oviposition by Ae. aegypti (Colton et al., 2003; Reiter, 2007) would increase the probability that ovipositing mosquitoes encounter a SLO or BLO. This suggests that mass trapping should be effective as a dengue control method despite the presence of competing containers, unless the number of competing containers is very numerous, and assuming there is effective coverage of target areas with the LOs. The oviposition rate in both the SLOs and the BLOs was comparable, with a mean of ∼6–13 eggs/lethal ovistrip. This suggests that the BLO is as effective as the SLO in attracting ovipositing mosquitoes. Traps that failed as a result of drying out, overturning or biodegradation had oviposition rates similar to wet traps in two out of three interventions, indicating that they had killed comparable numbers of mosquitoes before drying out.

Lethal ovistrip toxicity remained high despite the 4-week field exposure. In the SLO trial, despite a significant decrease in kill, 83% of the exposed mosquitoes were killed by exposed lethal ovistrips (Fig. 3). Our cage oviposition trial of BLO lethal ovistrips deployed for 4 weeks found a mean (±SE ) mortality of 74 ± 3%. In large cage choice oviposition trials, Williams et al. (2007) found that mortality averaged ∼80% over the 4-week exposure period of bifenthrin lethal ovistrips. Most BLO lethal ovistrips were in good condition after 4 weeks, although fungus and gelatinous films caused by breakdown of the BLO bucket did impact on many strips. Sithiprasasna et al. (2003) attributed the poor kill by LOs in Thailand in 1999 to fungal growth on the lethal ovistrips. Use of a fungicide could reduce fungal contamination, but may also reduce mosquito attraction and add to the complexity and cost of the operation.

The biodegradability of the BLOs appears to be satisfactory. Only 26% of the recovered LOs structurally failed, with most leaking as a result of chewing, probably by insects. The cockroach Periplaneta americana (L.) is common in urban areas of Cairns, and has been observed to cause damage similar to that shown in Fig. 2 when exposed to BLOs in our laboratory. Setting BLOs on porous (e.g. mulch, grass and soil) or solid (e.g. concrete, tile, stone) substrates did not significantly influence BLO failure. Nonetheless, 58% of BLOs were egg positive, with ∼2 traps/premise. BLO predation may vary, and trials should be conducted under local conditions to determine typical predation (BLO longevity) before adopting BLO for use. Furthermore, samples from BLO batches should be filled with water and observed for 2 days to ensure there are no structural faults with the batch.

Both SLOs and BLOs appear to be well suited for rapid mass trapping schemes for the control of dengue. Public acceptance, or at least lack of overt rejection, is probably high in north Queensland because setting the traps is less intrusive, and uses much less pesticide (∼1/1000th), compared with residual insecticide treatments inside homes (Ritchie, 2005). The new powers of the Queensland Health Act (2005) that, when a dengue surveillance and control programme is formally declared, enable health workers to set SLOs/BLOs and conduct larval control in the yard, even when residents are not present, has enabled QH to treat over 90% of premises in an intervention, with up to 50 premises treated by a two to three person team per day. This is about twice as fast as the insecticide spraying scheme previously employed (Ritchie, 2005). Both the SLOs and BLOs were subject to comparably high rates of mosquito oviposition. The lethal ovistrips maintained a lethality of 70–80% after 4-weeks field deployment, comparable with results obtained by Williams et al. (2007). This is also reflected in the significantly lower number of eggs laid on lethal versus control ovistrips.

In a recent forum paper, Morrison et al. (2008) supported the use of LOs for control of Ae. aegypti, and suggested they could be employed for the control of chikungunya and yellow fever viruses as well as dengue. We have shown here that our SLOs and BLOs are effective in attracting Ae. aegypti and appear to have public acceptance in this Australian situation. In a companion paper (Rapley et al., 2009), we describe the impact of the SLO and BLO interventions on local populations of Ae. aegypti. We are also studying the potential non-target impact of SLOs and BLOs, and establishing the longevity and mosquito productivity of BLOs that remain in the field.


We thank the Dengue Action Response Team and Avril Underwood of James Cook University for aiding in the field work. We thank Brian Montgomery, Ross Spark and Brad McCulloch of Queensland Health for their support. Finally, we thank the communities of Machans Beach and Parramatta Park for their cooperation. This work was funded by Australian National Health and Medical Research Council grant 379615 to SAR and RCR.